Connect your
entire stack.
Mode Agent works with every major AI provider, dev tool, and platform. No plugins, no adapters. Just connect and go.
AI Model Providers
Anthropic
Claude Opus, Sonnet, and Haiku. Best-in-class for strategy, writing, and complex reasoning.
OpenAI
GPT, o-series, and Codex models. Industry standards for code, chat, and reasoning.
Gemini Pro and Flash. Long-context models with strong multimodal capabilities.
xAI
Grok and Grok Code. Real-time knowledge and fast inference.
Mistral
Mistral Large, Small, and Codestral. European models with strong multilingual support.
Z.ai
GLM, GLM Air, and GLM Flash. Frontier open-weight models with strong agentic coding performance.
Moonshot AI
Kimi and Kimi Thinking. Trillion-parameter open-weight models built for agentic work and long context.
DeepSeek
DeepSeek Chat and DeepSeek Reasoner. High-performance reasoning models at competitive prices.
Alibaba Cloud
Qwen3, Qwen Coder, and Qwen VL. The largest open-weight family, with strong multilingual and coding performance.
Development tools
tmux
Native tmux integration for managing parallel agent sessions in your terminal.
Git & GitHub
Agents commit, branch, and create PRs. Full git workflow automation built in.
CLI & Shell
Run Mode Agent from any terminal. Pipe inputs, chain commands, and script agent workflows.
Your existing AI
Coding agents
Keep Claude Code, Codex, Cursor, or whatever agent you love. Mode's MCP server and CLI let it call Mode mid-session for models and tasks beyond its reach.
Model gateway
Point any tool that accepts a custom endpoint at Mode, and every Mode model shows up inside it. Billed to your Mode budget.
AI subscriptions
Where providers allow it, connect the AI subscriptions you already pay for and Mode routes eligible calls through them.
Specialized integrations
Embeddings
Voyage AI, Jina AI, and Cohere embeddings for semantic search and RAG pipelines.
Voice & Audio
ElevenLabs TTS, Deepgram and AssemblyAI STT for voice-enabled agent workflows.
Image & Video
Stability AI, Recraft, Ideogram for image generation. Runway and Luma for video.