Models

Open Llama instruction model for multilingual chat, reasoning, and coding

Cohere retrieval model for long-context chat and enterprise RAG workflows

Popular open Llama workhorse for multilingual chat, coding, and self-hosting

Flagship Mistral model for advanced reasoning, coding, and multilingual work

Tiny open Qwen code model for lightweight completion and on-device coding

Open coding-focused Qwen model for code generation, repair, and repository reasoning

Open-weight instruction model for adaptable chat and self-hosted production workloads

Flagship Mistral model for advanced reasoning, coding, and multilingual work

Mistral's larger vision model for document-heavy image understanding and chat

Open multilingual model optimized for generation across 23 languages

Open-weight instruction model for adaptable chat and self-hosted production workloads

Efficient open Mistral edge model for on-device chat and function calling

Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads

Compact open Llama base model for lightweight and on-device use

Small open Llama base model for lightweight text generation and self-hosting

Open multimodal Llama model for image understanding, captioning, and visual QA

Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants

Large open multimodal Llama model for image reasoning, document QA, and captioning

Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval

Small open Qwen2.5 instruction model for chat, structured output, and tool use

Mistral vision-language model for image understanding and multimodal chat

Cohere retrieval model for long-context chat and enterprise RAG workflows

Cohere's RAG workhorse for long-context enterprise search and tool use

Open Llama instruction model for multilingual chat, reasoning, and coding

Compact Nemotron model for efficient reasoning and deployable AI agents

Open Llama instruction model for multilingual chat, reasoning, and coding

Open Llama instruction model for multilingual chat, reasoning, and coding

Abacus.AI coding fine-tune of Llama 3.1 70B Instruct

Open Llama instruction model for multilingual chat, reasoning, and coding

Small open Gemma 2 instruction model for lightweight chat and on-device use

Multilingual Gemma 2 based embedding model for retrieval and semantic search

Flagship Mistral model for advanced reasoning, coding, and multilingual work

Llama 3.1-based safety classifier for moderating prompts and model responses

Compact open Llama model for lightweight chat, drafting, and self-hosting

Open Llama instruction model for multilingual chat, reasoning, and coding

Efficient Mistral-NVIDIA open model for multilingual chat and local deployment

Open Gemma instruction model for efficient chat and self-hosted deployments

Mistral code model for completions, refactors, and developer IDE workflows

Open Mistral code model for fill-in-the-middle and 80+ programming languages

Open 7B Mistral instruction model with function calling

Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue

Mistral model for multilingual chat, reasoning, and tool-assisted workflows

Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks

Lightweight multilingual cross-encoder that scores query-document relevance

Multilingual embedding model combining dense, sparse, and multi-vector retrieval

Open sparse mixture-of-experts Mistral instruction model

Granite 4.1 8B is a dense, decoder-only 8-billion-parameter language model from IBM, part of the Granite 4.1 family. It supports a 131K-token context window and is designed for enterprise tasks...

[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. At 14 billion...

Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,...