Cohere is a Toronto-founded enterprise AI company developing language models, document intelligence and the North platform for agents and automated workflows. Its portfolio combines private deployment options with model-specific licensing and commercial access routes.
Operating context
Cohere offers a choice between operating its models in a private environment and buying managed inference through Model Vault. With the latter, North’s application and control plane can run in the customer’s network while model requests go to Cohere-operated vault endpoints. Standard Vault provides logical isolation, with shared authentication and underlying compute components. Encrypted Vault adds confidential CPU/GPU execution and an attestation-verifying local OHTTP proxy or Python transport; its hosted demo proxy changes that trust boundary. Zero Data Retention must be enabled. Buyers size a model and performance tier against latency and throughput objectives: Fixed plans reserve instances, while Flex supports autoscaling and instance-hour overages. Encrypted Vault is a design-partner beta with a narrower documented model list, and neither Canadian hosting nor workload approval follows from the word sovereign. The Thales relationship remains a Canadian naval-support development route; it does not establish that Model Vault is deployed in fleet service.
Canadian footprint
Cohere was founded in Toronto in 2019; OpenText’s September 2026 announcement identifies Toronto and San Francisco headquarters and a Montreal office. Toronto is also a proposed headquarters for the combined Cohere/Aleph Alpha business, conditional on closing. Separately, ISED describes Cohere as the anchor tenant for CoreWeave’s AI compute expansion in Cambridge, Ontario, within a CAD 725 million project receiving up to CAD 240 million in federal contribution support. That compute project does not establish where a particular Model Vault endpoint runs; region availability needs confirmation for the selected service.
Toronto, Ontario is the published city-level context for Cohere. It does not imply a street address or exact facility location.
Cohere gives Canadian teams an enterprise AI route spanning private North workflows and managed model inference. Model Vault separates application operation from the GPU serving stack; its Encrypted beta adds verifiable protection of data in use for evaluation. The practical conversation is configuration-specific: selected model, Canadian hosting evidence, client verification, retention and workload acceptance. The Thales naval-support collaboration and planned OpenText integration provide distinct development and distribution routes; neither establishes operational Model Vault acceptance, and the Aleph Alpha combination remains conditional.
Interpretation of the public record. Not qualification, endorsement or evidence of buyer interest.
Cohere describes Command A+ as an open-weight mixture-of-experts model for private enterprise reasoning, retrieval-augmented generation, SQL and document analysis. Its Apache 2.0 licence differs from the non-commercial licence on North-Small-Translate weights.
Evidence of maturity
Released May 20, 2026; vendor-reported deployment and performance claims
Commercial availability
Apache 2.0 open weights; infrastructure and application validation remain the operator’s responsibility
Core features
218 billion total and 24 billion active parameters, according to Cohere
48 languages and multimodal document inputs
Vendor minimum deployment: two H100 GPUs or one B200
Private cloud, on-premises and air-gapped deployment options
Managed model inference and confidential computing
Read technical summary
Model Vault is Cohere’s managed inference service for running its models on dedicated serving infrastructure. Standard Vault provides logical isolation and encryption in transit and at rest; Encrypted Vault adds hardware-backed protection while data is processed and client-verifiable attestation. The encrypted offering is a design-partner beta with its own model list and access conditions. Applications can call vault endpoints directly, or run North in their own network while Cohere operates the inference layer.
Evidence of maturity
Standard Vault is commercially offered; Encrypted Vault is documented as a design-partner beta. Published architecture describes vendor controls rather than independent workload acceptance. No attained TRL is stated.
Commercial availability
Standard Embed/Rerank are self-serve; other models may require a waitlist. Encrypted beta is free for design partners through sales enrollment; general-availability pricing and Canadian region availability require confirmation.
Core features
Configurable performance tiers and 1–25 replicas per model, with request, latency, token and GPU monitoring
Standard Vault isolates serving components but shares authentication and underlying Kubernetes/compute resources; it does not provide encryption in use
North is Cohere’s enterprise platform for agents and automated workflows. Its July 2026 Automations release combines scheduled execution, branching and loops with model selection at each step, approval checkpoints and governance controls.
Evidence of maturity
Commercial platform; release capabilities reported by Cohere
Commercial availability
North Automations available to North customers from July 27, 2026
Core features
Scheduled workflows with loops and branching
Per-step model choice, review mode and version history
Human approval checkpoints, roles, permissions and token-use analytics
First-party integrations, MCP and SDK; on-premises or cloud deployment
North-Small-Translate-1.0 is Cohere’s mixture-of-experts text translation model for more than 50 languages. Its September 2026 release specifies 218 billion total and 25 billion active parameters, with 16k input and 16k output context.
Evidence of maturity
Released September 10, 2026; technical metrics are vendor-reported
Commercial availability
CC BY-NC 4.0 downloadable weights for non-commercial use; commercial enterprise access through RWS Language Weaver
Core features
More than 50 languages
218 billion total parameters and 25 billion active parameters
16k input and 16k output context
Minimum one B200 or two H100 GPUs at W4A4, according to Cohere
Parse converts document images, tables and text into reading-order Markdown for downstream retrieval and analysis. Cohere’s August 2026 release supports nine languages and provides table and image bounding boxes, with no per-text-element boxes or numeric chart-series extraction.
Evidence of maturity
Generally available from August 27, 2026; benchmarks are vendor-reported
Commercial availability
Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker; private deployment options
Core features
Reading-order Markdown from mixed document content
Nine-language support
Bounding boxes for tables and images; chart-series extraction excluded
Recorded applications
Document preparation for enterprise search and retrieval
Structured ingestion of reports and mixed-layout documents
On September 16, 2026, Cohere signed a definitive combination agreement with Aleph Alpha, subject to regulatory approvals, and OpenText announced a partnership to integrate North with its information-management and Aviator offering. OpenText expects integrated products in early 2027. These developments broaden the proposed sovereign AI and enterprise distribution footprint while leaving transaction closure and product delivery ahead.
North Automations is available to North customers, while Parse is generally available through Cohere and named cloud channels with private deployment options. Command A+ has Apache 2.0 open weights. North-Small-Translate-1.0 weights are restricted to non-commercial use under CC BY-NC 4.0, with enterprise translation offered through RWS Language Weaver.
Financing and disclosed activity
Cohere announced an EUR 500 million structured-financing commitment from companies of Schwarz Group in April 2026 to back an upcoming Series E and the planned Aleph Alpha combination. The cited announcement does not establish a closed financing round or cash received. Separately, the May 2026 PSPC briefing discusses up to CAD 240 million in sovereign AI compute support; this is not software procurement revenue.
Ownership and entity scope
Cohere and Aleph Alpha signed a definitive business-combination agreement on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; the proposed Toronto and Berlin headquarters and combined leadership remain contingent on completion.
Thales and Cohere announced joint development of agentic AI for Canadian naval and maritime in-service support in December 2025. The agreement establishes an integration and research route; it does not establish accepted fleet deployment.
AI education partnership
University of Waterloo
Waterloo and Cohere are developing an AI Transformation and Change Management certificate. Waterloo schedules its first cohort for January 2027 with 20 co-op positions, combining organizational problem discovery, implementation and change management.
Enterprise product and reseller partnership
OpenText
OpenText’s September 16, 2026 announcement describes integration of North agentic orchestration with OpenText information management and Aviator, with OpenText as a SOLEX strategic reseller. Integrated products are expected in early 2027.
Reviewed connections
Mission area · Reviewed connectionEvidence strength: Needs Review
The publicly described Enterprise generative AI offering may be relevant to monitoring or remote-operation use cases; this is a derived mapping for reviewer assessment.
Reviewed connections indicate possible relevance based on public evidence. They do not indicate procurement direction, eligibility, endorsement or customer interest.
Questions for a first conversation
01
Which model, region and hosting configuration meet the workload’s data boundary, latency and cost requirements?
North can use Cohere-managed Model Vault inference while its application runs in the customer network. Standard and Encrypted have different protection, model and access conditions; Fixed and Flex plans require workload sizing.
02
What attestation policy, client configuration, retention setting and acceptance evidence would govern an Encrypted Vault pilot?
The beta relies on verified CPU/GPU execution and a local OHTTP client path. ZDR is conditional, the hosted demo proxy changes trust, and hardware, metadata and service-availability risks remain. Canadian region and workload authorization must be established for the chosen deployment.
03
What naval-support workflow, evaluation evidence and acceptance authority would govern a Canadian pilot?
The Thales relationship establishes development intent. The public announcement does not identify an accepted operational configuration or its assurance evidence.
Original sources 26 linked records
These records support the organization, its owned capabilities and their reviewed connections. A source count is not a count of independent confirmations.
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
Supports: Profile · operating context
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
Supports: Profile · operating context
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
Cohere announced a definitive business combination agreement with Aleph Alpha on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; proposed Berlin and Toronto headquarters, more than 1,000 combined employees and leadership appointments are contingent on completion.
Supports: Profile · ownership
Cohere announced a definitive business combination agreement with Aleph Alpha on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; proposed Berlin and Toronto headquarters, more than 1,000 combined employees and leadership appointments are contingent on completion.
Public sources cited · Facts and assessments kept separate · Human review
Encrypted Vault combines confidential CPU virtual machines and attested NVIDIA GPUs, with OHTTP encryption into the protected environment
Local OHTTP proxy or conseel Python transport verifies attestation before sending requests; the hosted demonstration proxy introduces additional trust in Cohere
Zero Data Retention applies when enabled; OHTTP keys remain inside the attested environment
Standard Fixed plans reserve instances; Flex adds autoscaling and instance-hour overages against agreed latency objectives
Encrypted beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models; confirm the exact version, GPU and region with Cohere
Recorded applications
Managed inference for enterprise search, document processing and supervised agent workflows
Confidential-computing evaluation for sensitive inference workloads, subject to workload-specific assurance
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
Supports: Model Vault dedicated and encrypted AI inference · core features
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
Supports: Model Vault dedicated and encrypted AI inference · technical tags
OHTTP private keys are generated inside the TEE; attestation-gated peer sharing allows replicas to decrypt requests. Cohere states prompts/responses are not retained when ZDR is enabled; this is a configuration-dependent claim.
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Supports: Model Vault dedicated and encrypted AI inference · defence applications
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Supports: Model Vault dedicated and encrypted AI inference · summary
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Supports: Model Vault dedicated and encrypted AI inference · technicaldomainslugs
Customers run North application/control-plane components in their VPC/network and route inference to Cohere-managed vault endpoints. Standard and Encrypted types work with North, with attestation verification before encrypted inference.
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Supports: AI development partnership relationship
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Supports: AI development partnership relationship
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Supports: AI development partnership relationship
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Supports: AI development partnership relationship
Thales and Cohere announced a partnership on December 8, 2025 to develop agentic AI for naval and maritime in-service support in Canada, including joint research and secure AI integration associated with cortAIx. The release establishes a development partnership, not accepted Canadian Armed Forces fielding or a delivered operational configuration.
Source location:
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
Supports: Model Vault dedicated and encrypted AI inference · core features
A local OHTTP proxy container or conseel Python transport checks attestation before encrypting requests. The optional hosted demo proxy changes the trust model by requiring trust in Cohere at the proxy layer and is disabled by default.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
Cohere announced a definitive business combination agreement with Aleph Alpha on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; proposed Berlin and Toronto headquarters, more than 1,000 combined employees and leadership appointments are contingent on completion.
Supports: Profile · canadian footprint
Cohere announced a definitive business combination agreement with Aleph Alpha on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; proposed Berlin and Toronto headquarters, more than 1,000 combined employees and leadership appointments are contingent on completion.
Supports: Profile · canadian footprint
Cohere signed the definitive Aleph Alpha agreement on September 16, 2026, with closing expected later in the year subject to regulatory approvals. Proposed Toronto/Berlin headquarters and leadership appointments depend on closing; existing ownership/current-activity content remains supported.
Supports: Profile · current activity as of
Cohere announced a definitive business combination agreement with Aleph Alpha on September 16, 2026. Closing is expected later in 2026 subject to regulatory approvals; proposed Berlin and Toronto headquarters, more than 1,000 combined employees and leadership appointments are contingent on completion.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
Cohere signed the definitive Aleph Alpha agreement on September 16, 2026, with closing expected later in the year subject to regulatory approvals. Proposed Toronto/Berlin headquarters and leadership appointments depend on closing; existing ownership/current-activity content remains supported.
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Supports: Model Vault dedicated and encrypted AI inference · core features
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
Supports: Model Vault dedicated and encrypted AI inference · commercial availability
Encrypted Vault beta lists Embed, Rerank, Parse, Compass, North Mini Code, Transcribe and customized models. Model, GPU, capacity and region availability require dashboard or sales confirmation; the list does not establish Command family parity with Standard Vault.
OpenText describes Cohere as founded in Toronto in 2019, headquartered in Toronto and San Francisco, with a Montreal office among its international offices.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Supports: Profile · current activity
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Supports: Profile · description
OpenText describes Cohere as founded in Toronto in 2019, headquartered in Toronto and San Francisco, with a Montreal office among its international offices.
Supports: Profile · canadian footprint
OpenText describes Cohere as founded in Toronto in 2019, headquartered in Toronto and San Francisco, with a Montreal office among its international offices.
Supports: Enterprise product and reseller partnership relationship
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Supports: Enterprise product and reseller partnership relationship
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Supports: Enterprise product and reseller partnership relationship
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Supports: Enterprise product and reseller partnership relationship
OpenText and Cohere announced a strategic partnership on September 16, 2026 to combine OpenText information management and data context with North agentic orchestration in OpenText Aviator. Integrated products are expected in early 2027 and OpenText will be a SOLEX strategic reseller. Proposed deployment spans on-premises, private, public and sovereign clouds; the announcement is not evidence of present integrated-product delivery.
Cohere links enterprise inquiries to https://cohere.com/contact-sales from the North Automations announcement.
Supports: Profile · description
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: Profile · commercial status
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Cohere links enterprise inquiries to https://cohere.com/contact-sales from the North Automations announcement.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Cohere links enterprise inquiries to https://cohere.com/contact-sales from the North Automations announcement.
Supports: Profile · operating context
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · core features
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · maturity
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · core features
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · summary
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · technology readiness level
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · technicaldomainslugs
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · technical tags
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · name
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · capability type
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · commercial availability
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · defence applications
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · defence applications
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · technical tags
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · core features
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · core features
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Supports: North enterprise agents and workflow automation · technical tags
On July 27, 2026 Cohere made North Automations available to all North customers. North supports scheduled workflows, loops and branching, model choice by step, approval checkpoints, versioning and token-use analytics. North connects through first-party integrations, MCP or an SDK, and supports on-premises or cloud deployment with roles and permissions.
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Supports: Model Vault dedicated and encrypted AI inference · core features
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
Supports: Model Vault dedicated and encrypted AI inference · maturity
Standard Vault offers Fixed reserved instances or Flex baseline plus autoscaling with additional instance-hour billing, monthly/annual commitments and pause/resume. Tier selection follows latency/throughput SLOs; model/tier sizing matters more than a general no-rate-limit promise.
The official capability material supports the named Enterprise generative AI offering and the bounded operational description prepared for reviewer consideration.
Supports: Profile · canadian footprint
The official capability material supports the named Enterprise generative AI offering and the bounded operational description prepared for reviewer consideration.
Cohere's public evidence establishes its current role and offering, but it does not answer every access, configuration, eligibility, maturity or buyer question needed for a first business-development conversation.
The official public contact pathway for Cohere is https://cohere.com/.
Supports: North enterprise agents and workflow automation · summary
The official capability material supports the named Enterprise generative AI offering and the bounded operational description prepared for reviewer consideration.
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Supports: Profile · canadian footprint
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Supports: Profile · operating context
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Cohere describes controls supporting GDPR/HIPAA/SOC 2 obligations and asks customers to confirm available regions. This does not identify a Canadian Model Vault region or a Canadian defence workload authorization.
Standard Vault protects data in transit and at rest without hardware protection of data in use or remote attestation. Its architecture shares authentication and underlying Kubernetes/compute resources; Zero Data Retention applies when enabled.
Supports: Model Vault dedicated and encrypted AI inference · summary
Standard Vault protects data in transit and at rest without hardware protection of data in use or remote attestation. Its architecture shares authentication and underlying Kubernetes/compute resources; Zero Data Retention applies when enabled.
Supports: Model Vault dedicated and encrypted AI inference · core features
Standard Vault protects data in transit and at rest without hardware protection of data in use or remote attestation. Its architecture shares authentication and underlying Kubernetes/compute resources; Zero Data Retention applies when enabled.
On April 24, 2026 Cohere announced a EUR 500 million structured-financing commitment from companies of Schwarz Group intended to back an upcoming Series E and the planned Aleph Alpha combination. The announcement describes a commitment, not a closed round or cash received. The planned combined sovereign offering would use STACKIT cloud infrastructure.
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: Profile · commercial status
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · core features
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · defence applications
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · slug
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · commercial availability
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · technicaldomainslugs
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · name
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · summary
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · core features
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · maturity
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · technical tags
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · technical tags
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · core features
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · capability type
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · technology readiness level
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Supports: North-Small-Translate text translation · core features
North-Small-Translate-1.0 is a mixture-of-experts text translation model supporting more than 50 languages, with 218 billion total and 25 billion active parameters and 16k input/output context. Cohere specifies minimum hardware of one B200 or two H100s at W4A4. Downloadable weights use CC BY-NC 4.0 for non-commercial use; enterprise translation is offered through RWS Language Weaver.
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Profile · commercial status
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · defence applications
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · summary
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · technical tags
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · technical tags
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · capability type
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · commercial availability
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · technology readiness level
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · core features
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · defence applications
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · core features
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · name
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · technical tags
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · slug
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · core features
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · technicaldomainslugs
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · maturity
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Supports: Command A+ enterprise language model · core features
Cohere released Command A+ on May 20, 2026 as an Apache 2.0 open-weight mixture-of-experts model. Cohere reports 218 billion total and 24 billion active parameters, 48 languages and deployment on two H100s or one B200. It targets retrieval-augmented generation, SQL and document analysis in private cloud, on-premises or air-gapped environments; these deployment and performance statements are vendor claims.
Intel Trust Authority checks composite CPU/GPU evidence and registered policy; the client verifies its signed token and bound OHTTP key before sending data. Attestation identifies code, not whether that code is free of vulnerabilities. Cohere documents reproducible builds and remaining hardware, firmware and client trust.
Intel Trust Authority checks composite CPU/GPU evidence and registered policy; the client verifies its signed token and bound OHTTP key before sending data. Attestation identifies code, not whether that code is free of vulnerabilities. Cohere documents reproducible builds and remaining hardware, firmware and client trust.
Intel Trust Authority checks composite CPU/GPU evidence and registered policy; the client verifies its signed token and bound OHTTP key before sending data. Attestation identifies code, not whether that code is free of vulnerabilities. Cohere documents reproducible builds and remaining hardware, firmware and client trust.
Intel Trust Authority checks composite CPU/GPU evidence and registered policy; the client verifies its signed token and bound OHTTP key before sending data. Attestation identifies code, not whether that code is free of vulnerabilities. Cohere documents reproducible builds and remaining hardware, firmware and client trust.
Supports: Model Vault dedicated and encrypted AI inference · technical tags
Intel Trust Authority checks composite CPU/GPU evidence and registered policy; the client verifies its signed token and bound OHTTP key before sending data. Attestation identifies code, not whether that code is free of vulnerabilities. Cohere documents reproducible builds and remaining hardware, firmware and client trust.
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
Supports: Profile · operating context
The Thales-issued release carried by ENP/Acquiremedia describes a December 2025 agreement with Cohere for agentic AI development and naval in-service support in Canada. It establishes collaboration, not Model Vault use or accepted fleet deployment.
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Model Vault dedicated and encrypted AI inference · technical tags
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Model Vault dedicated and encrypted AI inference · summary
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Model Vault dedicated and encrypted AI inference · technicaldomainslugs
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Model Vault dedicated and encrypted AI inference · defence applications
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
Supports: Model Vault dedicated and encrypted AI inference · core features
Cohere states Encrypted Vault processes inference inside confidential CPU VMs and attested NVIDIA GPUs, with client OHTTP encryption terminating inside the TEE. Hardware/firmware, client and attestation remain trusted; side channels, traffic metadata and operator-controlled availability remain residual risks.
The May 5, 2026 PSPC briefing discusses up to CAD 240 million in Canadian Sovereign AI Compute Challenge support for Cohere and an August 2025 memorandum of understanding to explore government AI deployment. A memorandum and compute funding are not a government software purchase or operational acceptance.
ISED describes up to CAD 240 million toward a CAD 725 million Cohere project and an anchor-tenancy arrangement at a CoreWeave-built and operated AI facility in Cambridge, Ontario. This confirms Canadian compute expansion context, not the hosting region of a particular Model Vault.
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
Supports: Profile · operating context
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
Supports: Model Vault dedicated and encrypted AI inference · summary
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
Supports: Model Vault dedicated and encrypted AI inference · commercial availability
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
Supports: Model Vault dedicated and encrypted AI inference · technology readiness level
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
Supports: Model Vault dedicated and encrypted AI inference · maturity
Cohere documents Encrypted Vault as beta, free for design partners, with sales-led enrollment and full pricing to be shared as it matures. This does not establish generally available commercial encrypted service.
OpenText announces product integration and SOLEX reseller collaboration with Cohere, expecting client availability in early 2027. Its company description identifies Cohere as Toronto-founded in 2019, headquartered in Toronto and San Francisco with a Montreal office. Existing corporate and partner context remains supported; this is not Model Vault customer-acceptance evidence.
Supports: Profile · canadian footprint
OpenText announces product integration and SOLEX reseller collaboration with Cohere, expecting client availability in early 2027. Its company description identifies Cohere as Toronto-founded in 2019, headquartered in Toronto and San Francisco with a Montreal office. Existing corporate and partner context remains supported; this is not Model Vault customer-acceptance evidence.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · name
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · core features
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · maturity
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · capability type
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · core features
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · core features
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · summary
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Parse document intelligence · slug
Cohere Parse converts document images, tables and text to reading-order Markdown in nine languages. It is generally available through the Cohere API, Model Vault, Microsoft Foundry and AWS SageMaker, with private deployment options. The release limits bounding boxes to tables and images and excludes chart numeric-series extraction; benchmarks are vendor-reported.
Supports: Model Vault dedicated and encrypted AI inference · technical tags
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · capability type
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · summary
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · core features
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · name
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · slug
Cohere describes Model Vault as a managed single-tenant inference service with dedicated model-serving components, configurable model/performance tiers, 1–25 replicas per model and monitoring; it can serve applications directly or North.
Supports: Model Vault dedicated and encrypted AI inference · commercial availability
Standard Vault lists Command, North, Embed, Rerank, Parse, Compass, Transcribe and customized models. Embed/Rerank are self-serve; other models or bundles may need a waitlist. GPU and region availability are confirmed with Cohere.
Supports: Model Vault dedicated and encrypted AI inference · defence applications
Standard Vault lists Command, North, Embed, Rerank, Parse, Compass, Transcribe and customized models. Embed/Rerank are self-serve; other models or bundles may need a waitlist. GPU and region availability are confirmed with Cohere.
University of Waterloo describes a Cohere-supported AI Transformation and Change Management certificate and a first cohort scheduled for January 2027 with 20 co-op positions. The program combines organizational problem discovery, implementation and human-centred change management. This counterpart schedule is later than the fall 2026 timing described by Cohere.
Supports: AI education partnership relationship
University of Waterloo describes a Cohere-supported AI Transformation and Change Management certificate and a first cohort scheduled for January 2027 with 20 co-op positions. The program combines organizational problem discovery, implementation and human-centred change management. This counterpart schedule is later than the fall 2026 timing described by Cohere.
Supports: AI education partnership relationship
University of Waterloo describes a Cohere-supported AI Transformation and Change Management certificate and a first cohort scheduled for January 2027 with 20 co-op positions. The program combines organizational problem discovery, implementation and human-centred change management. This counterpart schedule is later than the fall 2026 timing described by Cohere.
Supports: AI education partnership relationship
University of Waterloo describes a Cohere-supported AI Transformation and Change Management certificate and a first cohort scheduled for January 2027 with 20 co-op positions. The program combines organizational problem discovery, implementation and human-centred change management. This counterpart schedule is later than the fall 2026 timing described by Cohere.
Continue exploring
Follow the strongest connections
Continue through related organizations, reviewed mission and Defence need connections, programme pathways, and explicitly linked intelligence. Similarity results describe shared areas of work, not partnerships or endorsements.
Organizations
Profiles that provide the strongest reviewed or discovery-led continuation.
