Command R7B (12-2024) Review
cohere/command-r7b-12-2024A small, fast Cohere text model for retrieval-based answers, tool use, agents, and complex reasoning.
128K tokens of context — about 96,000 words, on a scale that ends at 2M.
What it can do
- Takes text
- Gives text
Released 14 Dec 2024 · checked 3 Oct 2026.
About it
Command R7B (12-2024) is Cohere’s December 2024 text model. It accepts text and returns text, with a 128,000-token context window and outputs up to 4,000 tokens. The model is aimed at retrieval-augmented generation, tool use, agents, and other tasks that need complex reasoning. Input costs $0.0375 per million tokens; output costs $0.15.
Review
In a line
Cohere released Command R7B (12-2024), a small, fast model, on December 14, 2024. It is an update to Command R+ and works with text input and text output. See Cohere’s model list for related models.
What it's good at
The model is built for retrieval-augmented generation (RAG), where it uses supplied information while answering. It also suits tool use, agent tasks, and other work that calls for complex reasoning.
As a small, fast update to Command R+, it is intended for these tasks.
Price in plain terms
Input costs $0.0375 per million tokens. Output costs $0.15 per million tokens. Tokens are pieces of text counted for billing, with input covering what you send and output covering what the model writes.
Reading a page of about 500 words costs under 0.01 cents. A chat turn using about 1,000 input tokens and 500 output tokens costs about 0.011 cents. One thousand such turns cost about $0.11. Compare it with other low-cost language models.
Context and limits
The context window is 128,000 tokens, so one request can include a large amount of text for the model to consider. Its maximum output is 4,000 tokens in one answer.
This is a text-only model: the listed input and output formats are text, not images. No other limits or supported media are specified.
Ways to use it
Particular apps, chat tools, integrations, or API providers are not named.
Verdict
Choose Command R7B if you need an inexpensive text model for RAG, tool use, agents, or reasoning tasks and can work within a 4,000-token answer limit.
Its 128,000-token context window is useful when requests contain a lot of text.
Verdict
Command R7B (12-2024) fits buyers who want low token costs and a large text window for retrieval-based work, agent workflows, or tool-driven tasks. The 128,000-token context allows substantial material in one request, while the 4,000-token output cap keeps answers bounded. It is not a match for image-based work because only text input and output are specified. Buyers seeking a documented app, integration, or API route will need details beyond those provided here.
- 128,000-token context window for large text requests
- Low input rate of $0.0375 per million tokens
- Supports RAG, tool use, agents, and complex reasoning
- Maximum response length of 4,000 tokens
- Responses are capped at 4,000 tokens
- Specific apps, integrations, and API providers are not named
- Pick it ifRetrieval-based text tasks with large source material; Tool use, agents, and complex reasoning; High-volume chats where token cost matters
- Skip it ifYou need image input or image output; You require a documented app or integration; Your answers regularly need more than 4,000 tokens
Pricing
Input is $0.0375 per million tokens and output is $0.15 per million tokens. A 500-word page costs under 0.01 cents to read; 1,000 typical chat turns cost about $0.11.
Questions
What is Command R7B (12-2024)?
It is a Cohere text model released on December 14, 2024, and described as a small, fast update to Command R+.
What can Command R7B do?
It is suited to retrieval-augmented generation, tool use, agents, and tasks requiring complex reasoning.
How much does it cost?
Input costs $0.0375 per million tokens and output costs $0.15 per million tokens.
How much text can it read?
Its context window is 128,000 tokens.
How long can one answer be?
The maximum output is 4,000 tokens.
Does it handle images?
No image capability is listed; the specified input and output modality is text.