router
Model router for agentic systems. Routes every prompt to the right model in <50ms. Cut costs 40-70% with just an endpoint change.
Explore the three translation compatibility modes Off Shadow and Enforce in WorkWeave Router. Understand how each mode governs compatibility between AI providers and the internal model catalog.
How WorkWeave Router's Semantic Cache Determines Response EquivalenceDiscover how WorkWeave Router's semantic cache determines response equivalence using cosine similarity and configurable thresholds for efficient caching.
What Is the TTL for Session Pinning Entries in WorkWeave Router?Discover the default 30-minute TTL for session pinning entries in WorkWeave Router. Learn how to configure this setting using the ROUTER_SESSION_PIN_TTL environment variable for optimized performance.
How Session Pinning State Is Persisted and Invalidated in WorkWeave RouterDiscover how WorkWeave Router persists session pinning state in PostgreSQL and invalidates pins via TTL expiration error thresholds degenerate response detection loop detection and policy deadline violations.
Enterprise Custom Gateway Configuration in WorkWeave Router: A Deep DiveLearn how WorkWeave Router handles enterprise configurations with BYOK, routing traffic through private gateways by filtering requests against tenant-specific providers and model aliases.
Which AI Model Providers Are Supported by WorkWeave Router? A Complete Technical GuideWorkWeave Router supports 12 AI model providers like OpenAI, Gemini, and Bedrock via a unified architecture. Discover all compatible providers and integrate seamlessly.
How Provider Clients Are Registered and Managed in WorkWeave RouterLearn how WorkWeave Router registers and manages provider clients at startup. Discover its centralized mapping strategy for intelligent request routing and credential tracking.
Where Are Embedder Assets Stored in the WorkWeave Router Docker Image?Discover where embedder assets are stored in the WorkWeave Router Docker image. Learn how Go embed directive compiles assets into the router binary for read-only access via the artifacts variable.
What Embedder Backends Does WorkWeave Router Support?Discover the ONNX Runtime and stub backends supported by WorkWeave Router. Choose between local inference or efficient testing solutions for your projects.
How WorkWeave Router Artifacts Are Versioned and Selected for DeploymentDiscover how WorkWeave Router versions artifacts in discrete directories and selects them for deployment using the ROUTER_CLUSTER_VERSION environment variable or the latest symlink.
WorkWeave Router Artifact Bundles: Complete File Structure and LayoutUnderstand the WorkWeave Router artifact bundle file structure including centroids bin model registry metadata yaml and version specific JSON files for embedder data
How cluster.Scorer Determines Routing Decisions in the WorkWeave RouterLearn how the WorkWeave cluster.Scorer makes routing decisions by filtering candidates scoring models blending quality cost and speed to select the best option.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →