Cohere
Efficient

Command R7B

Compact

Smallest current Command R: 7B dense, single-GPU friendly, with the same RAG/tool focus as the larger family.

Command R7B is a efficient AI model from Cohere. It costs $0.0375 per million input tokens and $0.1500 per million output tokens (blended $0.0713/M), with a 128,000-token context window.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$0.0375/M
per million input tokens
OUTPUT
$0.1500/M
per million output tokens
BLENDED 70/30
$0.0713/M
unchanged since 3 May
CONTEXT
128,000
tokens

Hand-maintained from Cohere's published pricing; no check date on record.

What it is good at
  • Single-GPU
  • RAG-tuned
  • Cheap
Typical use cases
  • On-prem enterprise RAG
  • Cheap citation-grounded chat

Benchmarks

vs. best public score
MMLU65.2%
Multitask academic knowledge across 57 subjects. Source: Cohere (self-reported), MMLU, read 2026-11-10.
Python function synthesis from docstrings. Source: Cohere (self-reported), HumanEval (Python), read 2026-11-10.
LMArena Elo1115 Elo
Crowd-sourced head-to-head preference Elo rating. No source recorded yet.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

More from Cohere

See all 12 →

Frequently asked questions

How much does Command R7B cost?

Command R7B costs $0.0375 per million input tokens and $0.1500 per million output tokens, for a blended reference rate of $0.0713 per million tokens.

What is Command R7B's context window?

Command R7B supports up to 128,000 tokens of context in a single request.

What is Command R7B best for?

Command R7B is well suited to Single-GPU, RAG-tuned and Cheap.

Who makes Command R7B?

Command R7B is developed and served by Cohere. It was released in Dec 2024.

Terms used on this page