OpenFrog
Unified AI gateway for agent and developer workflows
Use one Base URL and one API key policy to route model traffic, control access, and standardize AI tooling across your team.
OpenFrog is designed for Codex, Claude Code, Gemini CLI, and other gateway-friendly clients.
Why OpenFrog?
A single control plane for model access, routing, usage governance, and AI agent integration.
Codex and CLI friendly
Bring agent capabilities into coding and shell-based workflows without rebuilding each client integration.
One API surface
Reuse OpenAI-style patterns so teams can reduce SDK drift, scattered environment variables, and duplicated setup.
Routing and operations
Keep upstream switching, quotas, and audit controls inside the gateway so operations stay centralized.
Documentation Areas
Quick Start
Start with the Base URL, API key, available models, and the first successful request.
CLI Setup
Configure Codex, Claude Code, Gemini CLI, and similar tools with a reusable gateway pattern.
Extensions and Troubleshooting
Centralize editor integrations, script-based access, and the most common debugging checks.
Current model families
OpenFrog documentation covers gateway setup and discovery language for GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GLM-5.2, Kimi K2.7 Code, DeepSeek V4 Pro, Qwen3.7 Max, MiniMax M2.7, Claude, Gemini, embeddings, rerank, image generation, and agent workflows.
OpenAI-compatible routes
Use familiar chat, responses, embeddings, image, and tool-oriented patterns through one gateway.
Agent clients
Prepare Codex, Claude Code, Gemini CLI, and extension workflows around one Base URL and key strategy.
Operational controls
Keep model access, billing, usage logs, quotas, and provider fallback in a centralized console.