cohere · Balanced
Cohere Command A (08/2025)
Cohere's first reasoning model designed for enterprise customer service and automation. 111B parameters with tool-use capabilities, supports 256K context and 23 languages including English, French, Spanish, Japanese, Arabic, and Hindi. Optimized for document processing, scheduling, data analysis, and more.
- Input
- Text
- Output
- Text
- Discovery tier
- Balanced
Pricing
USD per 1M tokens- Input
- $2.50
- Output
- $10.00
Reference prices as published by the model provider. API example
Specifications
- Context window
- 256K
- Max output
- 8.2K
- Average output (7 days)
- 371 tokens
- Longer than 38% of 468 models
Side by side
Cohere Command A (08/2025) compared
Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.
| Cohere Command A (08/2025)cohereThis model | Gemini 2.5 Pro Preview 0605GoogleRemove | |
|---|---|---|
| Provider | cohere | |
| Discovery tier | Balanced | Balanced |
| Added | Aug 22, 2025 | Jun 5, 2025 |
| Limits | ||
| Context window | 256K | 1.05MBest |
| Max output | 8.19K | 65.54KBest |
| Average output | 371 tokens | 727 tokens |
| Pricing, USD per 1M tokens | ||
| Input | $2.50 | $2.50 |
| Output | $10.00 | $10.00 |
| Cache read | — | $0.25 |
| Performance | ||
| Request success (24h) | — | 99.6% |
| Benchmark indexes | ||
| Intelligence Index | — | 16.1 |
| Coding Index | — | 33.3 |
| Math Index | — | 87.7 |
| Capabilities | ||
| Reasoning | No | Yes |
| Vision | No | Yes |
| Audio input | No | Yes |
| Reasoning effort | — | low, medium, high |
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
- Cohere Command A (08/2025)
- Gemini 2.5 Pro Preview 0605
Intelligence Index
- Cohere Command A (08/2025)—
- Gemini 2.5 Pro Preview 060516.1
Indexes
Coding Index
Composite of the coding evaluations
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 33.3
Math Index
Composite of the math evaluations
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 87.7
Agentic work
AutomationBench-AA
Workflow automation, penalized for guardrail violations
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 2.2%
AutomationBench-AA tasks completed
Workflows finished end to end without a guardrail violation
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 0.0%
AA-Briefcasev1.1
Agentic knowledge work, rated head to head
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 302 Elo
GDPval-AAv2.1
Economically valuable occupational tasks, rated head to head
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 455 Elo
Document reasoning
GDP.pdf
Professional reasoning over PDF documents, all-pass rate
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 10.2%
AA-LCRv1.1
Reasoning across long documents
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 69.0%
MLCR-AA
Long-context reasoning over medical records
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 12.2%
Reasoning and knowledge
Humanity’s Last Exam
Expert-level questions across academic fields
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 22.5%
GPQA Diamond
Graduate-level science questions
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 84.4%
MMLU-Pro
Multi-subject knowledge and reasoning
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 86.2%
IFBench
Following precise output instructions
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 48.7%
Coding
Terminal-Benchv4.0
Practical software engineering in a terminal
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 0.0%
SciCode
Scientific computing in Python
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 46.3%
LiveCodeBench
Recent competitive programming problems
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 80.1%
Math
AIME 2025
Competition mathematics, 2025 problems
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 87.7%
AIME
Competition mathematics
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 88.7%
MATH-500
Competition math problems
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 96.7%
Legacy evaluations
Terminal-Bench Hard
Earlier terminal task suite
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 26.5%
τ²-Bench Telecom
Tool use in a customer-support simulation
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 54.1%
AA-LCR (unversioned)
Long-context reasoning, earlier grading
- Cohere Command A (08/2025): —
- Gemini 2.5 Pro Preview 0605: 69.0%
Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis
- Cohere Command A (08/2025)
- Gemini 2.5 Pro Preview 0605
| Leaderboard | Cohere Command A (08/2025) | Gemini 2.5 Pro Preview 0605 |
|---|---|---|
| EQ-Bench 3 | Not in top 20 | #17 · 1523 |
Ranks among Noumen catalog models on EQ-Bench leaderboards. Only top-20 placements are listed.EQ-Bench
Developers
Use it through the API
Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "command-a-reasoning-08-2025",
"messages": [{ "role": "user", "content": "Explain this stack trace." }]
}'import os
from openai import OpenAI
client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
model="command-a-reasoning-08-2025",
messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)cohere
Related models
Other cohere models, starting with the same family.
- BalancedCompare