Introducing the New Auto and Auto Code Routers
OpenRouter released the Auto Router and Auto Code router. These tools evaluate factors like live pricing, provider health, and benchmarks to select an appropriate model based on governance policies.
46 releases · 1 this month
OpenRouter released the Auto Router and Auto Code router. These tools evaluate factors like live pricing, provider health, and benchmarks to select an appropriate model based on governance policies.
Admins can now configure organization-wide limits independently from workspace defaults. This allows for granular control over budgets and model access.
OpenRouter updated its provider routing algorithm to continuously monitor endpoint performance, optimizing speed, cost, and reliability for inference requests.
OpenRouter introduced automated in-flight screening of all inference requests to detect child sexual abuse material during request processing.
OpenRouter introduced Private Endpoints for Enterprise customers, allowing teams to route fine-tuned models and dedicated deployments through the standard OpenRouter API with organization-only visibility.
OpenRouter added ElevenLabs models to its catalog, offering nine Text-to-Speech models, including Eleven v4 and v4 Turbo, alongside two Speech-to-Text models like Scribe v2.
OpenRouter has introduced a Router Index to benchmark model routers based on quality, speed, and cost.
OpenRouter introduced a Security Center to help users manage and audit API keys. The tool identifies risky keys and allows users to lock access to specific networks.
OpenRouter launched its Batch API, enabling users to submit request batches in a single POST request and receive output within 24 hours at a 50% discount on per-token pricing.
OpenRouter launched US In-Region Routing via `us.openrouter.ai`, guaranteeing that API requests are decrypted and processed exclusively within United States infrastructure.
OpenRouter released the shell server tool and Files API in beta, giving models hosted Linux containers to execute shell commands, manage output, and upload or download files.
OpenRouter has launched the Analytics API and updated its dashboard to help teams track AI spend. Users can view costs per model, save custom charts, click bars to view underlying logs, and query usage data via the terminal.
OpenRouter has launched a new Auto router that uses market-based model decisions to route tasks. The router outperforms conventional task-based classifiers across a wide spectrum of tasks.
OpenRouter has released the ori CLI, allowing users to run supported harnesses with an optimized configuration out of the box after logging in once.
OpenRouter has launched Ori Eval, a tool that runs agents on custom prompts, checks the tools they call, and grades the answers to help developers choose the best model for their product.
OpenRouter has introduced a dedicated /api/v1/images endpoint for image generation models, using the same billing and keys as chat completions.
OpenRouter now allows users to define a taxonomy to tag generations by department, task type, or complexity, enabling filtered logs and grouped analytics.
OpenRouter has introduced an audio transcription API at /api/v1/audio/transcriptions, supporting base64 audio input and returning JSON text.
OpenRouter now supports chat, image generation, embeddings, and transcription through a single base URL and API key structure.
The OpenRouter MCP server allows coding agents to access the OpenRouter model catalog, benchmarks, and documentation directly from within an editor.
The new Unified Image API provides capability discovery across over 30 image models from 8 providers through a single endpoint.
OpenClaw now includes built-in support for OpenRouter, providing access to over 300 models with automatic failover.
The OpenAI Codex CLI can now be configured to route requests through OpenRouter for usage tracking and provider failover.
The new subagent tool allows frontier models to delegate self-contained tasks to smaller, faster worker models during generation.
The new Advisor tool enables smaller, cost-effective models to consult stronger models during generation for improved performance.
OpenRouter has released new speech and transcription APIs, Model Fusion, private models, and enterprise workspace controls, alongside 20 new model launches.
OpenRouter has introduced configurable security and governance tools, including budget enforcement, zero data retention, and prompt injection defense.
The OpenRouter Agent SDK now includes tools that allow agents to auto-resolve routine decisions while pausing for human input on high-stakes tasks. This implementation requires no loop-management code.
OpenRouter now enables tool-calling models to perform web searches and fetch page content. Users can choose from multiple search and fetch engines to integrate these capabilities into their workflows.
OpenRouter launched two new endpoints for audio transcription and speech synthesis. These APIs provide access to multiple providers under a single interface.
OpenRouter introduced a Response Caching header that enables caching for identical API requests. This feature reduces latency and eliminates costs for repeated requests.
Users can now create an OpenRouter account, generate an API key, and set up Stripe billing directly from the command line using Stripe Projects. This allows for automated account provisioning.
The OpenRouter Agent SDK provides a callModel function to build multi-turn agent workflows. It supports tool calls, stop conditions, and cost tracking across over 400 models.
OpenRouter now supports Workspaces, allowing users to organize projects into separate environments. Each workspace features independent API keys, routing defaults, guardrails, and observability settings.
OpenRouter launched a video generation API that provides access to top video models through a single interface.
OpenRouter has enabled Auto Exacto, an adaptive quality routing system, by default for requests involving tools. It re-evaluates providers every 5 minutes based on performance metrics.
OpenRouter has added model benchmarks to its pages, introduced a free model router, and improved cost transparency across providers.
OpenRouter released updates including auto-router customization, SDK skill loading, and general performance improvements for LLM access.
Response Healing is a new feature that automatically corrects malformed JSON responses from LLMs before they reach the application.
OpenRouter has updated the platform fee for its Bring Your Own Key (BYOK) service.
The unified model API and router.
| Release-notes page | Site | Feed |
|---|---|---|
| API Changelog - OpenRouter openrouter.ai | openrouter.ai | None |
| Ori Changelog openrouter.ai | openrouter.ai | None |