Audience
Developers and engineering teams using coding agents; platform engineers integrating model APIs.
About Weave Router
Weave Router is a model-routing service for coding-agent workflows. It analyzes incoming coding tasks and directs requests to selected language models, with routing informed by task complexity, cache-switching costs and task progress. Developers can connect coding tools including Claude Code, Codex, GitHub Copilot CLI, OpenCode and Pi using an installer, or configure Cursor manually. The service provides a compatible proxy for Anthropic, OpenAI and Gemini API formats, including streaming, tool calls and image inputs. Teams can use Weave’s hosted service or run the Apache-2.0-licensed proxy with their own provider credentials. Self-hosting gives teams control over deployment and provider charges; hosted plans are billed separately. Weave Router is intended for developers and engineering teams that want to route coding-agent model requests through managed or self-hosted infrastructure.