Command A
Better value: Qwen3.5-4B scores 49.3 for $0.060/M, 73× cheaper.
165 cheaper ranked models score higher: Gemini 4 Carbon (68.9, $4.00), GPT-6.1 Sol (66.6, $4.00), Gemini 4 Argon (65.6, $4.00) and 162 more.
As of 15 Nov 2026, Command A by Cohere scores 42.1 on the BenchLeader Index (611 ranked configurations), with its best results in human preference (52) and weakest in coding (32). API pricing: $2.50 per 1M input tokens and $10.00 per 1M output tokens ($4.38 blended at 3:1), list price on 15 Nov 2026. It streams about 44 tokens/s (Artificial Analysis). Released 13 Mar 2025.
Key facts (as of 15 Nov 2026)
- Command A API pricing: $2.50 per 1M input tokens and $10.00 per 1M output tokens ($4.38 blended at 3:1) (list price, 15 Nov 2026).
- Command A streams about 44 tokens/s (Artificial Analysis, 15 Nov 2026).
- Command A was released on 13 Mar 2025 (Cohere).
In detail. Command A is a Cohere open-weights model, released 13 Mar 2025. Its best configuration ranks #505 of 611 on the BenchLeader Index at 42.1 ±4.9, in the lower half. The ± is the point: 197 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (52) and lowest in coding (32). At $4.38 per million tokens blended it is among the most expensive ranked models. Output speed of 44 tokens per second puts it in the slowest quarter, with a first answer in 8.6 s. Last measured 11 Nov 2026.
- Blended price
- $4.38/M
- $2.50 in · $10.00 out
- via Artificial Analysis
- Output speed
- 44 tok/s
- measured by Artificial Analysis
- First answer
- 8.59 s
- first token 0.41 s
- Context
- 256k
- via models.dev
- Full answer
- 131 s
- last measured 13 Nov 2026
- Cost per task
- –
- Tokens per task
- –
- Released
- 13 Mar 2025
- Overall index42
- Reasoning40
- Coding32
- Agents & tools50
- Maths41
- Knowledge38
- Instruction following40
- Human preference52
- Long context34
Embed this rank
More badge styles[](https://www.benchleader.com/models/command-a)Paste it at the top of README.md, beside your other badges. The badge redraws with every data update and links back to this page.
How its index has moved
9 Sep 2026 to 15 Nov 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Metadata
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Openness Indexnot in index | 33.3#239 | Artificial Analysis |
Reasoning
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Hard Prompts | 1368#206 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 52.7%#422 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.0%#571 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 48.5%#121 | Vals AI | |
| Kagi LLM Benchmark | 28.8%#126 | Kagi LLM Benchmark | |
| CritPtnot in index | 0.0%#337 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Coding | 1389#212 | Arena (LMArena) | |
| LiveCodeBench | 35.1%#135 | Vals AI | |
| SWE-bench (Vals)not in index | 7.8%#87 | Vals AI | |
| LiveCodeBench (AA)not in index | 28.7%#210 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench Hard | 0.8%#354 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 15.2%#366 | Artificial Analysis | |
| BFCL Overall | 46.5%#28 | Berkeley Function Calling Leaderboard | |
| Terminal-Bench 2.0 (Vals)not in index | 2.3%#65 | Vals AI |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AIME (Vals) | 13.3%#85 | Vals AI | |
| MGSM | 85.7%#65 | Vals AI | |
| MATH 500not in index | 76.2%#43 | Vals AI | |
| Arena Maths | 1310#225 | Arena (LMArena) | |
| AIME 2025 (AA)not in index | 13.0%#216 | Artificial Analysis |
Knowledge
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 36.5%#305 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1342#207 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1354#204 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1336#190 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1360#193 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1370#194 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR v1.1 | 21.3%#451 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2not in index | 7.0#508 | Artificial Analysis | |
| AA Progress Index v1not in index | 158.1#542 | Artificial Analysis |
API pricing: where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Cohere | 53 tok/s | 0.37 s | $2.50 | $10.00 | 256k | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0040 | – | 15.4 s |
| Summarize a 30-page report | 12,000 / 600 | $0.036 | – | 22.2 s |
| Code edit | 6,000 / 1,500 | $0.030 | – | 42.7 s |
| Agentic coding session | 60,000 / 4,000 | $0.190 | – | 1.7 min |
| Structured extraction | 2,000 / 200 | $0.0070 | – | 13.1 s |
See also
Embed or cite
Show where Command A ranks on a model card or README. The badge redraws itself with every daily update, carries the data date and links back to this page.
[](https://www.benchleader.com/models/command-a)Cite this
The ranking changes daily, so a reference carries the data date. APA style:
BenchLeader. (2026, November 15). Command A on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 15, 2026, from https://www.benchleader.com/models/command-a
BibTeX:
@misc{benchleader_command_a_20261115,
author = {{BenchLeader}},
title = {{Command A on the BenchLeader Index (method v1.4)}},
year = {2026},
month = nov,
howpublished = {\url{https://www.benchleader.com/models/command-a}},
note = {Data as of 2026-11-15}
}Every badge style, the embeddable top-10 widget and how to cite the Index
Questions people ask
- What are Command A's benchmark scores?
- As of 15 Nov 2026, Command A scores 42.1 ±4.9 on the BenchLeader Index, strongest in human preference (52) and weakest in coding (32). That rests on independent benchmark results only, and every score on this page links to its publisher.
- How does Command A rank?
- Command A is #505 of 611 ranked configurations on 15 Nov 2026. 197 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Fable 5.2 (max effort) at 72.7.
- How much does Command A cost per 1M tokens?
- As of 15 Nov 2026, Command A costs $2.50 per 1M input tokens and $10.00 per 1M output tokens, $4.38 blended (three input tokens per output token).
- When was Command A released?
- Cohere released Command A on 13 Mar 2025.
Data as of 15 Nov 2026. Compare with another model.
Cite as: BenchLeader, “Command A: benchmarks, pricing, speed and rank”, data as of 15 Nov 2026, https://www.benchleader.com/models/command-a.