Vendor · Language models · Specialist models
Cohere has 23 models in the directory: 19 language models and 4 specialist models.
Cohere develops foundation models and AI products for business applications.
In the directory: 19 language models and 4 specialist models. 19 of them with open weights that can run locally.
Every model there ever was. Models with a profile lead to every benchmark, price and source.
| North-Small-Translate-1.0 | Sep 10, 2026 | Open weights | n/a | n/a | n/a | n/a | below 1 |
| North Mini Code | Jun 9, 2026 | Active | n/a | n/a | 68 | 256K | 35 |
| Command A Plus | May 20, 2026 | Active | below 1 | n/a | 150 | 128K | 32 |
| Tiny Aya Earth | Feb 17, 2026 | Open weights | n/a | n/a | n/a | n/a | 41 |
| Tiny Aya Fire | Feb 17, 2026 | Open weights | n/a | n/a | n/a | n/a | 29 |
| Tiny Aya Global | Feb 17, 2026 | Open weights | n/a | n/a | n/a | n/a | 45 |
| Tiny Aya Water | Feb 17, 2026 | Open weights | n/a | n/a | n/a | n/a | 29 |
| Command A Translate | Aug 28, 2025 | Active | n/a | n/a | n/a | n/a | 20 |
| Command A Reasoning | Aug 21, 2025 | Active | n/a | n/a | n/a | 256K | 22 |
| Command A Vision | Jul 31, 2025 | Active | n/a | n/a | n/a | 128K | 27 |
| Aya Vision 32B | Mar 4, 2025 | Active | n/a | n/a | n/a | 16K | 18 |
| Aya Vision 8B | Mar 4, 2025 | Active | n/a | n/a | n/a | 16K | 23 |
| Command A | March 2025 | Active | below 1 | $2.50 / $10 | 17 | 256K | 29 |
| Command R7B Arabic | Feb 27, 2025 | Active | n/a | n/a | n/a | 128K | 28 |
| Command R7B | December 2024 | Active | n/a | $0.038 / $0.15 | n/a | 128K | 40 |
| Aya Expanse 32B | Oct 24, 2024 | Active | n/a | n/a | n/a | 128K | 33 |
| Aya Expanse 8B | Oct 24, 2024 | Active | n/a | n/a | n/a | 8K | 38 |
| Command R+ | April 2024 | Deprecated | n/a | n/a | n/a | 128K | 28 |
| Command R | March 2024 | Deprecated | n/a | n/a | n/a | 128K | 49 |
| Cohere Rerank 5 Fast | Oct 21, 2026 | Active | n/a | below 1 |
| Cohere Rerank 5 Pro | Oct 21, 2026 | Active | n/a | below 1 |
| Cohere Embed v5 Fast | Sep 30, 2026 | Active | n/a | below 1 |
| Cohere Embed v5 Pro | Sep 30, 2026 | Active | n/a | below 1 |
Specialist model · Oct 21, 2026
Rerank 5 Fast is the smaller, throughput-oriented text reranker. Its API orders documents and returns calibrated relevance scores; Cohere lists about 500 million parameters and a limit of 32,768 tokens for the query plus each individual document.
Specialist model · Oct 21, 2026
Rerank 5 Pro orders text documents by relevance to a search query. Cohere lists about 2 billion parameters and a limit of 32,768 tokens for the query plus each individual document. Context Focus can return relevant text spans.
Specialist model · Sep 30, 2026
Cohere Embed v5 Fast shares a vector space with Embed v5 Pro, so documents indexed with Pro can be searched through the lower-latency, higher-throughput Fast variant.
Specialist model · Sep 30, 2026
Cohere Embed v5 Pro is optimized for retrieval quality. Cohere reports improvements over Embed 4 on visually rich documents, financial filings, PDFs, code, and multilingual retrieval.
Language model · Sep 10, 2026
North-Small-Translate-1.0 focuses Cohere Labs’ MoE architecture on translation across more than 50 languages. It activates 25 billion of its 218 billion parameters. Translation takes priority over general chat, with context and maximum output limited to 16K tokens each.
Language model · Jun 9, 2026
North Mini Code is Cohere's purpose-built open-weight checkpoint for code generation, agentic software engineering, and terminal tasks. The 30-billion-parameter MoE activates 3 billion parameters and was trained with tool use for coding agents.
Language model · May 20, 2026
Command A Plus extends the Command A family into a sparse mixture-of-experts model with image input. Cohere describes it as its first MoE model and names agentic tasks, reasoning, translation, and support for 48 languages. The model has 218 billion total parameters and 25 billion active parameters.
Language model · Feb 17, 2026
Tiny Aya Earth focuses on languages across Africa and West Asia. Cohere Labs offers it as a regional alternative to the balanced Global variant. The checkpoint has 3.35 billion parameters and belongs to the Tiny Aya family covering more than 70 languages. Its local weights use CC BY-NC 4.0 with additional Cohere terms for non-commercial use.
Language model · Feb 17, 2026
Tiny Aya Fire directs Cohere Labs’ compact multilingual model toward South Asia. Like Global, it has 3.35 billion parameters, but with a regional focus. It belongs to the Tiny Aya family covering more than 70 languages. Cohere releases the local weights for non-commercial use under CC BY-NC 4.0 with additional usage terms.
Language model · Feb 17, 2026
Tiny Aya Global balances its multilingual focus across world regions. Cohere Labs offers it alongside three region-specialized Tiny Aya models. The family covers more than 70 languages, and Global contains 3.35 billion parameters.
Language model · Feb 17, 2026
Tiny Aya Water prioritizes languages from Asia-Pacific and Europe. This 3.35B checkpoint complements Cohere Labs’ Tiny Aya family, while Earth focuses on Africa and West Asia and Fire on South Asia. The weights can run locally and use CC BY-NC 4.0 with additional Cohere terms for non-commercial use.
Language model · Aug 28, 2025
Command A Translate is a Command A specialist built for machine translation. Cohere names 23 languages and positions the checkpoint around translation work rather than general agent conversations.
9 more current models are in the model list above.
| Vendor | Models | Model types | Best language model | Top 10 | Open weights | New in 2026 |
|---|---|---|---|---|---|---|
| Anthropic | 33 | 33 language models | Claude Opus 5.5 (Rank 1 to 3) | 4 | 0 | 13 |
| OpenAI | 63 | 46 language models, 4 image models, 2 video models, 9 audio models, 2 specialist models | GPT-6 Astra (Rank 4 to 8) | 2 | 4 | 27 |
| 66 | 43 language models, 4 image models, 6 video models, 11 audio models, 2 specialist models | Gemini 4 Carbon (Rank 1 to 5) | 2 | 19 | 26 | |
| Meta | 32 | 29 language models, 1 image model, 1 video model, 1 specialist model | Muse Wonder Full (Rank 6 to 9) | 2 | 26 | 7 |
| DeepSeek | 23 | 23 language models | DeepSeek V4.1 Flash 1017 (Rank 13 to 20) | 0 | 23 | 9 |
| Moonshot AI | 12 | 12 language models | Kimi K3.1 Pro (Rank 10 to 13) | 0 | 10 | 7 |
| xAI | 16 | 12 language models, 2 image models, 2 video models | Grok 4.7 (Rank 18 to 24) | 0 | 1 | 12 |
| Cohere | 23 | 19 language models, 4 specialist models | not ranked yet | 0 | 19 | 11 |
By release date. Grey: retired models.
11 releases since January 2026, by release date.
Prices per 1M tokens for the current language models, according to Cohere or the provider linked in each model profile.
| Current model | Input | Output | Context |
|---|---|---|---|
| Command A | $2.50 | $10 | 256K |
| Command R7B | $0.038 | $0.15 | 128K |
The directory lists 23 Cohere models: 19 language models and 4 specialist models. It counts every model Cohere ever released, including retired ones. 21 are currently available, the newest being Cohere Rerank 5 Fast, Cohere Rerank 5 Pro and Cohere Embed v5 Fast.
Cohere Rerank 5 Fast, released on Oct 21, 2026. The timeline below shows every release by date.
Through the API per 1M tokens (input / output): Command R7B $0.038 / $0.15. The prices of all 2 current language models are in the price table.
Yes, 19 of the 23 models have open weights. Each model profile lists how much memory it needs.
Command A Plus with 150 tokens per second according to Artificial Analysis. A typical model manages 98.
Up to 256,000 tokens (≈ 380 A4 pages), which 3 of the current models reach, including North Mini Code. Every model's value is in the model list.
Data as of Nov 1, 2026