# ARouter

## Docs

- [Count Tokens](https://docs.arouter.ai/en/api-reference/anthropic/count-tokens.md): Anthropic-compatible token counting endpoint.
- [Create Message](https://docs.arouter.ai/en/api-reference/anthropic/create-message.md): Anthropic-compatible Messages API endpoint.
- [Create Message Batch](https://docs.arouter.ai/en/api-reference/anthropic/create-message-batch.md): Anthropic-compatible Message Batches API endpoint.
- [Get Balance](https://docs.arouter.ai/en/api-reference/billing/get-balance.md): Retrieve the current prepaid credit balance for your ARouter account.
- [List Transactions](https://docs.arouter.ai/en/api-reference/billing/list-transactions.md): List recent billing transactions, including usage charges and credit purchases.
- [List Endpoints for a Model](https://docs.arouter.ai/en/api-reference/endpoints/list-endpoints.md): Retrieve all available provider endpoints for a specific model, including pricing, context length, and capabilities.
- [Generate Content](https://docs.arouter.ai/en/api-reference/gemini/generate-content.md): Gemini-compatible generateContent endpoint.
- [List Models (Gemini)](https://docs.arouter.ai/en/api-reference/gemini/list-models.md): Gemini-compatible model listing endpoint.
- [Stream Generate Content](https://docs.arouter.ai/en/api-reference/gemini/stream-generate-content.md): Gemini-compatible streaming generateContent endpoint.
- [Get Generation](https://docs.arouter.ai/en/api-reference/generations/get-generation.md): Retrieve cost, token usage, and metadata for a specific completed generation.
- [API Reference](https://docs.arouter.ai/en/api-reference/introduction.md): Complete reference for the ARouter API. Request and response schemas are normalized across all providers so you only need to learn one interface.
- [Create API Key](https://docs.arouter.ai/en/api-reference/keys/create-key.md): Create a new API key with optional provider/model restrictions, rate limits, and expiry.
- [Delete API Key](https://docs.arouter.ai/en/api-reference/keys/delete-key.md): Delete an API key by its hash. Deleted keys are immediately invalidated.
- [List API Keys](https://docs.arouter.ai/en/api-reference/keys/list-keys.md): List all API keys managed by the current management key.
- [Update API Key](https://docs.arouter.ai/en/api-reference/keys/update-key.md): Update an existing API key's name, status, spending limits, or access restrictions. Management key required.
- [Create Chat Completion](https://docs.arouter.ai/en/api-reference/openai/create-chat-completion.md): OpenAI-compatible chat completions endpoint with multi-provider routing.
- [Create Embedding](https://docs.arouter.ai/en/api-reference/openai/create-embedding.md): OpenAI-compatible embeddings endpoint.
- [List Models](https://docs.arouter.ai/en/api-reference/openai/list-models.md): List all models available to the current API key.
- [Parameters](https://docs.arouter.ai/en/api-reference/parameters.md): Sampling parameters shape the token generation process of the model. ARouter passes these through to the upstream provider. If the chosen model doesn't support a parameter, it is silently ignored.
- [List Providers](https://docs.arouter.ai/en/api-reference/providers/list-providers.md): Retrieve the list of all AI providers available through ARouter.
- [Proxy Provider Request](https://docs.arouter.ai/en/api-reference/proxy/proxy-provider-request.md): Proxy a native provider request through ARouter using your ARouter API key.
- [Authentication](https://docs.arouter.ai/en/authentication.md): Four ways to authenticate with the ARouter API — pick the one that matches your use case.
- [FAQ](https://docs.arouter.ai/en/faq.md): Frequently asked questions about ARouter — authentication, billing, models, rate limits, and more.
- [Activity Export](https://docs.arouter.ai/en/guides/administration/activity-export.md): Export your ARouter usage data as CSV or PDF reports.
- [API Key Rotation](https://docs.arouter.ai/en/guides/administration/api-key-rotation.md): Best practices for rotating ARouter API keys with zero downtime. Keep your application secure by regularly cycling credentials.
- [Organization Management](https://docs.arouter.ai/en/guides/administration/organization-management.md): Manage teams and shared resources with ARouter organizations.
- [Usage Accounting](https://docs.arouter.ai/en/guides/administration/usage-accounting.md): How ARouter tracks and reports token usage, cost, and timing for every request.
- [User Tracking](https://docs.arouter.ai/en/guides/administration/user-tracking.md): Track API usage by individual users in your application using the user field. Useful for per-user analytics, cost allocation, and abuse detection.
- [Latency and Performance](https://docs.arouter.ai/en/guides/best-practices/latency-and-performance.md): Understanding ARouter's performance characteristics and best practices for optimal response times.
- [Reasoning Tokens](https://docs.arouter.ai/en/guides/best-practices/reasoning-tokens.md): Understanding and working with reasoning (thinking) tokens in models like DeepSeek R1, Claude Extended Thinking, and o4-mini.
- [Uptime Optimization](https://docs.arouter.ai/en/guides/best-practices/uptime-optimization.md): How ARouter tracks provider availability and maximizes uptime for your applications.
- [Billing & Credits](https://docs.arouter.ai/en/guides/billing-and-credits.md): Understand how billing works, check your balance, and view transaction history.
- [BYOK: Bring Your Own Key](https://docs.arouter.ai/en/guides/byok.md): Use your own LLM provider API keys with ARouter for direct control over rate limits, costs, and provider access.
- [Frameworks & Integrations](https://docs.arouter.ai/en/guides/community/frameworks-and-integrations.md): Use ARouter with popular AI frameworks and SDKs — zero code changes required beyond base_url and api_key.
- [Enterprise Quickstart](https://docs.arouter.ai/en/guides/enterprise-quickstart.md): Get your organization up and running with ARouter: organizations, API key management, security controls, and observability.
- [Error Handling](https://docs.arouter.ai/en/guides/error-handling.md): Understand ARouter error responses and how to handle them in your application.
- [Audio](https://docs.arouter.ai/en/guides/features/audio.md): Send audio files as input and receive spoken audio as output. ARouter supports audio transcription, translation, text-to-speech, and multimodal audio models.
- [Guardrails](https://docs.arouter.ai/en/guides/features/guardrails.md): Enforce organizational policies on every API request. Guardrails control which models can be used, set spending limits, and configure data retention — all from a central place.
- [Image Generation](https://docs.arouter.ai/en/guides/features/image-generation.md): Generate images from text prompts using models like DALL-E 3, Stable Diffusion, and Flux. ARouter routes image generation requests through a single unified endpoint.
- [Message Transforms](https://docs.arouter.ai/en/guides/features/message-transforms.md): Automatically compress prompts that exceed a model's context window using the context-compression plugin.
- [Model Variants](https://docs.arouter.ai/en/guides/features/model-variants.md): Append suffixes to any model ID to control routing behavior — optimize for speed, cost, reasoning depth, or extended context.
- [Multimodal](https://docs.arouter.ai/en/guides/features/multimodal.md): Send images, PDFs, and audio alongside text. Generate images and spoken audio. ARouter supports multimodal models from OpenAI, Anthropic, and Google.
- [Plugins](https://docs.arouter.ai/en/guides/features/plugins-overview.md): Extend any model's capabilities with ARouter plugins. Enable web search, context compression, auto-routing, and more via the plugins array.
- [Prompt Caching](https://docs.arouter.ai/en/guides/features/prompt-caching.md): Reduce costs and latency by caching repeated prompt prefixes. Supported by OpenAI, Anthropic, DeepSeek, and Google Gemini.
- [Response Healing](https://docs.arouter.ai/en/guides/features/response-healing.md): ARouter automatically repairs malformed JSON responses when using json_schema or json_object response formats. No changes needed in your application.
- [Datetime Tool](https://docs.arouter.ai/en/guides/features/server-tools/datetime.md): Inject the current date and time into model context. The datetime Server Tool gives models accurate time awareness without system prompt hacks.
- [Server Tools](https://docs.arouter.ai/en/guides/features/server-tools/overview.md): Give models access to real-time capabilities like web search and datetime. Server Tools are executed by ARouter on behalf of the model, with no additional client-side logic required.
- [Web Search Tool](https://docs.arouter.ai/en/guides/features/server-tools/web-search.md): Give models access to real-time web search. ARouter executes the search and injects structured results with citations — no client-side code required.
- [Service Tiers](https://docs.arouter.ai/en/guides/features/service-tiers.md): Control the tradeoff between cost and latency using the service_tier parameter. ARouter passes service tier preferences to providers that support it.
- [Structured Outputs](https://docs.arouter.ai/en/guides/features/structured-outputs.md): Force models to output valid JSON matching an exact schema. Use json_schema mode for guaranteed schema compliance.
- [Tool Calling](https://docs.arouter.ai/en/guides/features/tool-calling.md): Enable your AI to call functions and take actions in the world. Tool calling lets models interact with external APIs, databases, and services.
- [Web Search](https://docs.arouter.ai/en/guides/features/web-search.md): Ground any model's responses with real-time web data. Use the web plugin or :online suffix — or migrate to Server Tools for model-driven search.
- [Zero Data Retention (ZDR)](https://docs.arouter.ai/en/guides/features/zdr.md): Route requests exclusively to providers that have signed Zero Data Retention agreements. ZDR ensures your prompts and completions are never stored by the upstream provider.
- [Zero Completion Insurance](https://docs.arouter.ai/en/guides/features/zero-completion-insurance.md): ARouter does not charge for requests where the model returns zero output tokens due to an error or provider-side failure. Enabled by default.
- [For Providers](https://docs.arouter.ai/en/guides/for-providers.md): Integrate your models and endpoints with ARouter to reach developers using the unified API.
- [Key Management](https://docs.arouter.ai/en/guides/key-management.md): Create and manage API keys with provider restrictions, rate limits, spending limits, and expiry.
- [Rate Limits](https://docs.arouter.ai/en/guides/limits.md): ARouter API rate limits, credit-based quotas, and how to check your current limits.
- [Bring Your Own Key (BYOK)](https://docs.arouter.ai/en/guides/overview/byok.md): Use your own provider API keys through ARouter. BYOK routes requests directly to providers using your credentials — no ARouter markup on token costs.
- [Management API Keys](https://docs.arouter.ai/en/guides/overview/management-api-keys.md): Management API keys are separate credentials used exclusively for ARouter admin operations — creating, listing, and deleting API keys. They cannot be used for LLM inference.
- [Data Collection](https://docs.arouter.ai/en/guides/privacy/data-collection.md): What data ARouter collects and how it's used.
- [Provider Logging](https://docs.arouter.ai/en/guides/privacy/provider-logging.md): How upstream AI providers handle logging and data retention when you use ARouter.
- [Provider Proxy](https://docs.arouter.ai/en/guides/provider-proxy.md): Directly proxy requests to any supported provider while using your ARouter API key. No need to manage multiple provider keys.
- [App Attribution](https://docs.arouter.ai/en/guides/request-attribution.md): Identify your application in API requests using HTTP headers. Power your Dashboard analytics and app-level usage breakdowns.
- [Auto Exacto](https://docs.arouter.ai/en/guides/routing/auto-exacto.md): ARouter automatically re-ranks providers by tool-calling quality for requests that include tools. Disable it when you need predictable cost-based routing.
- [Model Fallbacks](https://docs.arouter.ai/en/guides/routing/model-fallbacks.md): Specify an ordered list of fallback models. If the primary model fails or is unavailable, ARouter automatically retries with the next model in the list.
- [Streaming](https://docs.arouter.ai/en/guides/streaming.md): Enable real-time token-by-token streaming with Server-Sent Events (SSE). ARouter supports streaming responses for all major providers.
- [x402 Agent Payments](https://docs.arouter.ai/en/guides/x402-payments.md): Let AI agents with crypto wallets automatically top up credits and authenticate — no registration, no dashboard, fully autonomous.
- [Introduction](https://docs.arouter.ai/en/introduction.md): One API key, every LLM provider. Route, manage, and observe all your LLM traffic through a single gateway.
- [Model Routing](https://docs.arouter.ai/en/model-routing.md): How ARouter routes requests to the right provider using the provider/model format, `auto`, and ordered candidate model lists.
- [Models](https://docs.arouter.ai/en/models.md): Query available models, compare capabilities, and browse model metadata via the Models API.
- [Principles](https://docs.arouter.ai/en/principles.md): ARouter's core values and mission: unified, open, and reliable AI access for every developer.
- [Provider Routing](https://docs.arouter.ai/en/provider-routing.md): ARouter routes requests to the best available provider automatically. Customize routing with the provider object — sort by price, throughput, or latency; filter by quantization; enforce data policies.
- [Supported Providers](https://docs.arouter.ai/en/providers.md): All LLM providers available through ARouter, with their supported models and endpoints.
- [Quick Start](https://docs.arouter.ai/en/quickstart.md): Start using ARouter in under 2 minutes. No code changes needed — just swap base_url and api_key.
- [cURL](https://docs.arouter.ai/en/sdks/curl.md): Quick testing and scripting with cURL.
- [SDK Downloads](https://docs.arouter.ai/en/sdks/downloads.md): Download official ARouter SDKs to integrate with your applications.
- [Go](https://docs.arouter.ai/en/sdks/go.md): Use ARouter with the official Go SDK.
- [Node.js / TypeScript](https://docs.arouter.ai/en/sdks/node.md): Use ARouter with the OpenAI, Anthropic, or ARouter Node.js SDK.
- [SDK Overview](https://docs.arouter.ai/en/sdks/overview.md): ARouter works with every major LLM SDK. Zero new dependencies needed.
- [Python](https://docs.arouter.ai/en/sdks/python.md): Use ARouter with the OpenAI, Anthropic, or Gemini Python SDK.

## OpenAPI Specs

- [openapi](https://docs.arouter.ai/api-reference/openapi.yaml)

## Optional

- [GitHub](https://github.com/arouter-ai)