LLM Reference

Command R

Released
2024-04-04
Last refreshed
2026-06-15
Status
Researched 90d ago
DeprecatedOpen weightsCommercial use: non-commercialCodingRAGLong contextClassificationJSON / Tool use

Command R is a legacy integration reference; keep it only while you identify a current replacement.

Use it for

  • Teams maintaining an existing integration
  • Workloads that can use a 128k context window
  • Buyers comparing 4 tracked provider routes

Do not use it for

  • New production launches
  • Vision or document-understanding workloads
Specifications
Family
Command R
Released
2024-04-04
Context
128k
Parameters
104B*
Architecture
Decoder Only
Specialization
general
Openness
Open weights
License
CC-BY-NC-4.0Commercial use: non-commercial
Weights
Unknown
Code
Unknown
Training
Fine-tuned
Created by

Empowering developers with advanced language AI.

Toronto, Ontario, Canada
Founded 2022
Website
Pricing
Output / 1M
$0.600
Input / 1M
$0.150

Cheapest of 6 routes · OpenRouter

About

Command R is a 35-billion parameter large language model developed by Cohere, tailored for enterprise-scale AI applications. It is optimized for long-context tasks such as retrieval-augmented generation (RAG) and tool use, emphasizing high performance coupled with strong accuracy. This model excels in reasoning, summarization, and question answering across ten key languages. Its use spans various enterprise applications, automating complex tasks that involve interacting with systems like CRMs, APIs, and databases. The model can process up to 128,000 tokens and integrates seamlessly with Cohere's Embed and Rerank models, enhancing accuracy and efficiency in large-scale tasks. Additionally, a more powerful iteration, Command R+ with 104 billion parameters, offers even greater capabilities.

Command R is an open-weight model. The structured metadata tracks a 128k-token context window and structured outputs. This page tracks provider routes through AWS Bedrock, Cohere API, Microsoft Foundry, and 3 more, with the cheapest tracked route listed at $0.15 input and $0.6 output per 1M tokens. Headline tracked benchmarks include Google-Proof Q&A 53.4, HellaSwag 90.8, and HumanEval 77.8.

Top use-case fit: coding, agents, and build tasks

Coding

1 relevant benchmark in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Long context

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare all 6

Compare API pricing across 4 providers for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MRoute
OpenRouter$0.150$0.600
Serverless
AWS Bedrock$0.500$1.50
Serverless
Cohere API$0.500$1.50
Serverless
Microsoft Foundry$0.500$1.50
Serverless

Available via routers & gateways(7)

Capabilities

Structured Outputs

Benchmark peer barsfor Coding

Benchmark scores(4)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionEvaluationSource
Google-Proof Q&A53.4diamondObserved 2026-03-06research
HellaSwag90.810-shotObserved 2026-03-06research
HumanEval77.8pass@1Observed 2026-03-06Source
Massive Multitask Language Understanding80.25-shotObserved 2026-03-06Source

Migration checks

No linked migration route is available for this model yet.

Compare Command R with other models

Frequently asked questions

What is the context window of Command R?

Command R has a context window of 128k tokens.

How much does Command R cost?

Command R pricing ranges from $0.15/1M to $0.5/1M input tokens depending on the provider.

When was Command R released?

Command R was released on 2024-04-04.

Which providers offer Command R?

Command R is available from 6 providers: AWS Bedrock, Cohere API, Microsoft Foundry, OCI Generative AI, OpenRouter, DeepInfra.

What benchmarks has Command R been tested on?

Command R has been evaluated on 4 benchmarks, including Google-Proof Q&A, HellaSwag, HumanEval, Massive Multitask Language Understanding.