Skip to content
← Models
Open weightsCohere Labs/10 trustMetrics-checked
Command R7B logo

Command R7B

7B multilingual RAG and tool-use model for research use.

Visit official page →

About

Cohere Labs' 7B research release built for retrieval-augmented generation and tool use, trained on 23 languages with 128K context and multi-step agentic capabilities.

It punches above its size on enterprise RAG and code tasks. The catch is the license: CC-BY-NC, so this is for research and prototyping only — not for commercial products.

Specs

Context window

128K tokens

Pricing

Free weights (CC-BY-NC, non-commercial); hosted API priced separately

Free quota

None (self-host for research)

API access

Cohere API, HuggingFace (gated, non-commercial)

Weights

Open weights

Released

Not disclosed

Benchmarks

Not publicly disclosed in the sources reviewed.

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

agentsdataautomation

Trust score — 6/10

/10 trustMetrics-checked

Scored 6/10 (metrics tier, metrics-v1). Evidence: Widely used for RAG prototyping; 23-language coverage is a real differentiator; Cohere Labs research release; commercial development happens on the closed Command line; Weights available but CC-BY-NC — non-commercial only, gated download. Manual review: 128K context and license terms verified on the Hugging Face model card. License: CC-BY-NC with Cohere acceptable use policy (non-commercial).

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
7.0/10
Maintenance · recency, releases, issue load
6.0/10
Openness · license + docs
5.0/10
Manual review · sanity read + red-flag check
7.0/10
  • · Widely used for RAG prototyping; 23-language coverage is a real differentiator
  • · Cohere Labs research release; commercial development happens on the closed Command line
  • · Weights available but CC-BY-NC — non-commercial only, gated download
  • · manual review: 128K context and license terms verified on the Hugging Face model card

Sources