Ctrl K
Back to models

Cohere

Command A+

  • Released
  • Open weights
  • Apache 2.0
  • Free Cohere API · limits apply
Compare
Intelligence Index

Reasoning setting

13.1#62 of 83measured models
Measured · Artificial Analysis v4.3.2
Context window
192K tok

Performance profile

ReasoningRank #60 of 107
76.1%
GPQA Diamond
CodingRank #71 of 94
38.5%
SciCode

Capabilities

Context window

Modalities

Measured setting

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
Command A+
Cost per task (USD) · log scale
Command A+Reasoning
Intelligence Index
13.1
Cost / task
$0.00
Evaluation time
2.0m
Rate-limited free API

Pricing & access

Cohere API access is free within rate limits. Paid provider routes, Model Vault and local hardware costs are separate.

No verified provider quote is listed in this snapshot.

Free Cohere API · limits apply (opens in a new tab)

Run locally

GGUF
Context
Estimated memory

~139.4 GiB

RAM + VRAM

Estimate assumptions

One text chat at the selected context, with a 16-bit cache and app allowance. Operating system and image processing are extra; optimized caches can use less.

192 GiB RAMwithout GPU offload

Memory budgetOne text chat
Weights
125.8 GiB
Context cache
1 GiB16-bit cache
App allowance
12.6 GiB
Download
135.1 GB4 files · keep together
Disk space
~154 GB
Available software
Ollama (opens in a new tab)LM Studio (opens in a new tab)Jan (opens in a new tab)KoboldCpp (opens in a new tab)llama.cpp (opens in a new tab)
Get GGUF (opens in a new tab)
Set up in Ollama · Q4_K_M · 8K context

Save the following as Modelfile. The context matches the estimate above.

FROM hf.co/bartowski/command-a-plus-05-2026-GGUF:Q4_K_M
PARAMETER num_ctx 8192

Then run these commands in your terminal:

ollama create llmcurrent-local -f Modelfile
ollama run llmcurrent-local

Ollama import instructions (opens in a new tab) · For LM Studio or Jan, open the app’s model search, enter bartowski/command-a-plus-05-2026-GGUF, choose Q4_K_M, and set context to 8,192 tokens in the app’s model settings. Software links above open each app’s setup guide.

Image support unverified (opens in a new tab)vLLM setup (opens in a new tab)

Benchmarks

Full leaderboardSource (opens in a new tab)

Humanity’s Last Exam

Frontier reasoning

Command A+ · Reasoning#83 of 126

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  5. 80Mistral Medium 3.5High13.8%
  6. 81EXAONE 4.5 33BReasoning12.9%
  7. 82Qwen3 Next 80B A3B (Reasoning)Reasoning12.6%
  8. 83Command A+Reasoning12.0%
  9. 84Nova 2.0 LiteHigh11.6%
  10. 85NVIDIA Nemotron 3 Nano 30B A3BReasoning11.4%
  11. 86Granite 4.2 30BReasoning11.2%

Peers use their best measured result on this benchmark.

Alternatives

Each model at its highest-scoring measured setting · Intelligence Index v4.3.2 and its evaluation cost per task.