Models for Coding

StarCoder2 is an open family of code-focused large language models from the BigCode ecosystem designed for code generation, completion, and developer-assistance workflows.

Mistral Devstral is an open-weight coding-focused model from Mistral AI designed for agentic software engineering tasks such as navigating repositories, editing code, and solving GitHub-style issues.

Mistral Codestral is a code-focused large language model from Mistral AI designed for code generation, completion, fill-in-the-middle editing, and developer-assistant workflows across many programming languages.

Qwen3-Coder is Alibaba Qwen’s code-focused large language model family designed for agentic software engineering, long-context code understanding, tool use, and automated coding workflows.

Qwen2.5-Coder is Alibaba Qwen's open-weight family of code-focused large language models for code generation, completion, reasoning, debugging, and developer-agent workflows.

DeepSeek-Coder-V2 is an open-weight mixture-of-experts coding language model family from DeepSeek designed for code generation, code understanding, debugging, and long-context software engineering tasks.

Google Gemini 2.5 Pro is Google DeepMind’s advanced reasoning-focused multimodal model with strong coding, long-context, and agentic software engineering capabilities.

Anthropic Claude 4 is a family of frontier AI models, including Claude Opus 4 and Claude Sonnet 4, optimized for advanced coding, agentic workflows, reasoning, and software engineering automation.

Anthropic Claude 3.7 Sonnet is a hybrid reasoning and coding-focused large language model designed for software engineering, agentic workflows, and complex problem solving.

Anthropic Claude 3.5 Sonnet is a high-performance frontier AI model optimized for reasoning, coding, text generation, and multimodal analysis with strong speed-cost efficiency.

OpenAI codex-1 is a software-engineering-focused AI model powering OpenAI Codex, designed to autonomously understand codebases, edit code, run tests, and produce implementation-ready patches.

OpenAI GPT-4.1 is a high-capability API model optimized for coding, instruction following, and long-context software engineering workflows.

Open-weight Chinese AI models combining mixture-of-experts architecture with advanced reasoning capabilities. They offer competitive coding performance against proprietary models at lower cost.

OpenAI's top-tier reasoning model optimized for complex problem-solving, code generation, and technical analysis with demonstrated strong performance on advanced coding benchmarks.

Anthropic's latest coding-focused model designed for complex software engineering tasks including multi-file refactoring, agentic workflows, and deep code understanding across large codebases.