Skip to content

AI model by Cohere

Cohere Command A+ for business

Cohere's enterprise flagship for multimodal, multilingual agent tasks, with 48 languages, tool calling, JSON output and open weights under Apache 2.0.

  • Reviewed on September 24, 2026
  • Cohere

Capability tiersRelative, not benchmarks

Reasoning High
Coding Medium
Vision Medium
Speed Medium
Price level Medium
Context size High
Tool calling Very high
Structured output High

Key facts about Cohere Command A+

Model ID at review
command-a-plus-05-2026
Provider
Cohere
Main category
General purpose
Open weights
Yes, Apache 2.0
Last reviewed
September 24, 2026

Fit

Where it fits and where it does not

Good at

  • Enterprise agent tasks with tools.

  • Many languages, 48 at review time.

  • Image input with text.

  • Running on modest hardware for its class.

  • Open weights under Apache 2.0.

Not the right choice for

  • Inputs over its 128K token window.

  • Audio or video input.

  • The hardest reasoning tasks, where top models lead.

  • Adoption without testing, since it is recent.

Use cases

Business use cases we would use it for

Multilingual support

Assistants that answer customers in many languages. See translation and localization.

Enterprise search assistants

Answering from company documents, paired with Cohere Embed and Rerank.

Private deployment

Running the open weights in your own environment.

Our notes

When we would choose it

Command A+ was released by Cohere in May 2026 and is now its flagship. Cohere positions it for complex multimodal, multilingual agentic enterprise tasks on minimal hardware. It supports 48 languages, image input, tool calling and JSON output, with a 128K token input window and up to 64K output tokens. It is a mixture-of-experts model, and an open-weight version is published under Apache 2.0.

Cohere's focus has always been enterprise use: search, retrieval and assistants for business data. Command A+ fits naturally with Cohere Embed and Cohere Rerank in a knowledge assistant built on one provider.

The open weights make it one of the few enterprise-focused models you can run in your own environment. For companies with strict data rules, that is a real advantage.

The older Command A from 2025 is still live, with a 256K window but text only. Cohere also lists specialized versions for reasoning, vision and translation. We did not verify pricing at review time, so the price level here is an estimate.

We would include Command A+ in a comparison when languages beyond English matter, or when open weights and enterprise support are both needed. See RAG knowledge assistants.

Before you commit

Things to check before you commit

  • Context

    The input window is 128K tokens with up to 64K output.

  • Pricing

    We did not verify API pricing at review time.

  • Self-hosting

    Check the hardware needs of the published open-weight version.

  • Older models

    Command A from 2025 is still live. Compare both.

Alternatives

Models to compare it with

Cohere Embed v4

Cohere's multimodal embedding model for enterprise search, embedding text, images and mixed documents, with flexible vector sizes and long inputs.

Cohere Rerank 4

Cohere's multilingual rerank models, which sort search results by relevance, in a best-quality Pro version and a low-latency Fast version.

Mistral Large 3

Mistral's open-weight general-purpose flagship under Apache 2.0, with image input, tool calling, structured outputs and a 256K token window.

Keep exploring

FAQ

Questions people ask us

Have a question that is not here? Ask us directly.

Start a project

Not sure which model fits? Ask us to evaluate your use case.

Send a short brief. We reply with questions, a suggested plan and an estimate you can compare with other offers.

Your privacy choices

We use necessary storage to run this site. With your permission we also use Google Analytics to see which pages help people, and load maps from Google. You can change this at any time. Read the cookie policy.