Cohere (BYOK)BYOK

Command A+ is Cohere's first Mixture-of-Experts model, with vision input, reasoning, and agentic tool use. It has 25B active parameters, supports 48 languages, and offers a 128K-token context window with up to 64K tokens of output. It is released open-weight.

Command A is Cohere's flagship model for agentic, multilingual, and enterprise use cases. It is optimized for tool use, retrieval-augmented generation, and complex reasoning, supports 23 languages, and offers a 256K-token context window. Released open-weight for research use.

Command R+ (08-2024) is Cohere's update of Command R+ with roughly 50% higher throughput and 25% lower latencies than the previous version. It is a text model with a 128K-token context window, streaming, and function calling.

Embed 5 Fast is the lower-latency tier of Cohere's Embed 5 family, at about 1B parameters. It turns text into 2048-dimension vectors across more than 100 languages with a 128K-token context. It shares one vector space with Embed 5 Pro.

Embed 5 Pro is the higher-accuracy tier of Cohere's Embed 5 family, at about 2.4B parameters. It turns text into 2048-dimension vectors across more than 100 languages with a 128K-token context. It shares one vector space with Embed 5 Fast.

North Mini Code is Cohere's first agentic coding model, a sparse mixture-of-experts with 30B total and 3B active parameters. It has a 256K-token context window with up to 64K tokens of output and interleaved reasoning with tool use. It is open-weight under the Apache 2.0 license.