# AUTO-GENERATED — DO NOT EDIT MANUALLY # Source: scripts/docs/generate_llms_txt.py # Regenerated on every deploy # Auriko > Intelligent LLM routing API with OpenAI-compatible interface. Route requests > across multiple AI providers to optimize for cost, latency, and reliability. Auriko provides a unified API for accessing LLMs from OpenAI, Anthropic, Google, Meta, Mistral, and more. Change 2 lines of code to start saving 30-70% on LLM costs. ## Get Started - [Introduction](https://docs.auriko.ai/introduction.md): Route LLM requests across providers for lower cost and higher reliability. - [Quickstart](https://docs.auriko.ai/quickstart.md): Send your first request in minutes - [OpenAI Compatibility](https://docs.auriko.ai/openai-compatibility.md): Minimal changes to your existing OpenAI code ## Guides - [Streaming](https://docs.auriko.ai/guides/streaming.md): Receive chat completion tokens as they're generated. - [Tool Calling](https://docs.auriko.ai/guides/tool-calling.md): Let LLMs call your functions to interact with external systems - [Structured Output](https://docs.auriko.ai/guides/structured-output.md): Get JSON responses that conform to a schema - [Vision](https://docs.auriko.ai/guides/vision.md): Analyze images alongside text in chat completions - [Image generation](https://docs.auriko.ai/guides/image-generation.md): Generate images with Gemini models in chat completions - [Routing Options](https://docs.auriko.ai/guides/routing-options.md): Optimize for cost, latency, or throughput - [Cost Optimization](https://docs.auriko.ai/guides/cost-optimization.md): Reduce inference costs with cache-aware routing - [Error Handling](https://docs.auriko.ai/guides/error-handling.md): Handle errors with retries, fallbacks, and typed exceptions - [Prompt Caching](https://docs.auriko.ai/guides/prompt-caching.md): Reduce costs and latency with automatic prompt caching - [Extensions and Thinking](https://docs.auriko.ai/guides/extensions-and-thinking.md): Control reasoning effort and pass provider-specific parameters through Auriko's normalized API - [Advanced Routing](https://docs.auriko.ai/guides/advanced-routing.md): Fine-tune routing with suffix shortcuts, multi-model requests, quality constraints, parameter filtering, custom weights, and data policies ## Response API - [Overview](https://docs.auriko.ai/response-api/overview.md): Create model responses using the OpenAI Response API format through Auriko - [Streaming](https://docs.auriko.ai/response-api/streaming.md): Stream Response API events as they're generated - [Tool Calling](https://docs.auriko.ai/response-api/tool-calling.md): Call functions using the Response API's input/output item format - [Structured Output](https://docs.auriko.ai/response-api/structured-output.md): Get JSON responses conforming to a schema using the Response API - [Reasoning](https://docs.auriko.ai/response-api/reasoning.md): Control reasoning effort and access reasoning summaries in Response API output - [Routing and Extensions](https://docs.auriko.ai/response-api/routing-extensions.md): Use multi-model routing, routing options, and provider extensions with the Response API ## Frameworks - [LangChain](https://docs.auriko.ai/frameworks/langchain.md): Use LangChain with Auriko's routing and cost optimization - [Vercel AI SDK](https://docs.auriko.ai/frameworks/vercel-ai-sdk.md): Use the Vercel AI SDK with Auriko's routing and cost optimization - [OpenAI Agents SDK](https://docs.auriko.ai/frameworks/openai-agents-sdk.md): Use the OpenAI Agents SDK with Auriko's routing and cost optimization - [Claude Agent SDK](https://docs.auriko.ai/frameworks/claude-agent-sdk.md): Use Claude Code and the Claude Agent SDK with Auriko's routing and cost optimization - [Google ADK](https://docs.auriko.ai/frameworks/google-adk.md): Use Google's Agent Development Kit with Auriko's routing and cost optimization - [CrewAI](https://docs.auriko.ai/frameworks/crewai.md): Use CrewAI with Auriko's routing and cost optimization - [LlamaIndex](https://docs.auriko.ai/frameworks/llamaindex.md): Use LlamaIndex with Auriko's routing and cost optimization ## Integrations - [Claude Code](https://docs.auriko.ai/integrations/claude-code.md): Use Auriko as your LLM provider in Claude Code - [Codex](https://docs.auriko.ai/integrations/codex.md): Use Auriko as your LLM provider in Codex - [Hermes Agent](https://docs.auriko.ai/integrations/hermes-agent.md): Use Auriko as your LLM provider in Hermes Agent - [OpenClaw](https://docs.auriko.ai/integrations/openclaw.md): Use Auriko as your LLM provider in OpenClaw - [OpenCode](https://docs.auriko.ai/integrations/opencode.md): Use Auriko as your LLM provider in OpenCode - [Kilo Code](https://docs.auriko.ai/integrations/kilo-code.md): Use Auriko as your LLM provider in Kilo Code ## Specifications - [Supported Parameters](https://docs.auriko.ai/contract/supported-parameters.md): Request parameters accepted by the Auriko Chat Completions API - [Error Codes](https://docs.auriko.ai/contract/error-codes.md): Machine-readable error codes returned by the Auriko API - [batch_only](https://docs.auriko.ai/errors/batch_only.md): HTTP 400 — Model is batch-only and cannot serve real-time requests. - [budget_exhausted](https://docs.auriko.ai/errors/budget_exhausted.md): HTTP 429 — Account or project budget reached. - [byok_keys_required](https://docs.auriko.ai/errors/byok_keys_required.md): HTTP 400 — BYOK keys required but none are configured. - [client_disconnected](https://docs.auriko.ai/errors/client_disconnected.md): Internal observability code — the client closed the connection before the streamed response completed. - [content_filtered](https://docs.auriko.ai/errors/content_filtered.md): HTTP 400 — The upstream provider's content filter rejects the request. - [context_length_exceeded](https://docs.auriko.ai/errors/context_length_exceeded.md): HTTP 400 — Token count exceeds the model's context window. - [cost_constraint_exceeded](https://docs.auriko.ai/errors/cost_constraint_exceeded.md): HTTP 400 — No provider meets the cost constraint. - [duplicate_resource](https://docs.auriko.ai/errors/duplicate_resource.md): HTTP 409 — Resource with this identifier already exists. - [expired_api_key](https://docs.auriko.ai/errors/expired_api_key.md): HTTP 401 — API key has expired. - [feature_disabled](https://docs.auriko.ai/errors/feature_disabled.md): HTTP 403 — Feature is not available for this account or plan. - [field_immutable](https://docs.auriko.ai/errors/field_immutable.md): HTTP 400 — Field cannot be modified after resource creation. - [hosted_tool_not_supported](https://docs.auriko.ai/errors/hosted_tool_not_supported.md): HTTP 400 — Model does not support the requested hosted tool on available providers. - [idempotency_conflict](https://docs.auriko.ai/errors/idempotency_conflict.md): HTTP 409 — Idempotency key reused with a different payload. - [idempotency_replay_unavailable](https://docs.auriko.ai/errors/idempotency_replay_unavailable.md): HTTP 409 — Original request completed but replay data was not captured. - [input_requires_responses_endpoint](https://docs.auriko.ai/errors/input_requires_responses_endpoint.md): HTTP 400 — Request input contains types served only via the native Response API, and no provider for the model offers it. - [insufficient_permissions](https://docs.auriko.ai/errors/insufficient_permissions.md): HTTP 403 — Caller authenticated but lacks the required scope. - [insufficient_quota](https://docs.auriko.ai/errors/insufficient_quota.md): HTTP 429 — No quota remaining on the current plan. - [internal_error](https://docs.auriko.ai/errors/internal_error.md): HTTP 500 — Unexpected server error. - [invalid_api_key](https://docs.auriko.ai/errors/invalid_api_key.md): HTTP 401 — API key is missing, malformed, revoked, or unrecognized. - [invalid_parameter_value](https://docs.auriko.ai/errors/invalid_parameter_value.md): HTTP 400 — Field present but value is outside the allowed range or set. - [invalid_recovery_code](https://docs.auriko.ai/errors/invalid_recovery_code.md): HTTP 401 — MFA recovery code is invalid or already used. - [invalid_request](https://docs.auriko.ai/errors/invalid_request.md): HTTP 400 — Generic malformed or invalid request. - [json_mode_not_supported](https://docs.auriko.ai/errors/json_mode_not_supported.md): HTTP 400 — Model does not support JSON mode. - [latency_constraint_exceeded](https://docs.auriko.ai/errors/latency_constraint_exceeded.md): HTTP 400 — No provider meets the latency constraint. - [method_not_allowed](https://docs.auriko.ai/errors/method_not_allowed.md): HTTP 405 — Request method not allowed for this endpoint. - [mfa_required](https://docs.auriko.ai/errors/mfa_required.md): HTTP 403 — Action requires multi-factor authentication step-up. - [mfa_verification_failed](https://docs.auriko.ai/errors/mfa_verification_failed.md): HTTP 401 — TOTP verification failed or code expired. - [missing_required_parameter](https://docs.auriko.ai/errors/missing_required_parameter.md): HTTP 400 — Required field is absent from the request. - [model_not_found](https://docs.auriko.ai/errors/model_not_found.md): HTTP 404 — Requested model ID isn't in the catalog. - [model_unavailable](https://docs.auriko.ai/errors/model_unavailable.md): HTTP 503 — The requested model is temporarily unavailable. - [no_compatible_endpoint](https://docs.auriko.ai/errors/no_compatible_endpoint.md): HTTP 400 — No provider supports the required API format for this model. - [no_provider_available](https://docs.auriko.ai/errors/no_provider_available.md): HTTP 503 — No provider can serve this request right now. - [no_responses_endpoint](https://docs.auriko.ai/errors/no_responses_endpoint.md): HTTP 400 — Request requires the Response API but no provider supports it for this model. - [non_streaming_not_supported](https://docs.auriko.ai/errors/non_streaming_not_supported.md): HTTP 400 — Model requires streaming. - [operation_not_allowed](https://docs.auriko.ai/errors/operation_not_allowed.md): HTTP 400 — Operation not permitted for this resource state or role. - [payload_too_large](https://docs.auriko.ai/errors/payload_too_large.md): HTTP 413 — Request body exceeds the size limit. - [platform_keys_unavailable](https://docs.auriko.ai/errors/platform_keys_unavailable.md): HTTP 400 — No platform keys available for this model. - [provider_blocked](https://docs.auriko.ai/errors/provider_blocked.md): HTTP 400 — All providers for this model are in the blocklist. - [provider_not_in_allowlist](https://docs.auriko.ai/errors/provider_not_in_allowlist.md): HTTP 400 — No provider in the allowlist supports this model. - [rate_limit_exceeded](https://docs.auriko.ai/errors/rate_limit_exceeded.md): HTTP 429 — Requests-per-minute cap hit. - [reasoning_not_supported](https://docs.auriko.ai/errors/reasoning_not_supported.md): HTTP 400 — Model does not support reasoning/extended thinking. - [required_params_not_supported](https://docs.auriko.ai/errors/required_params_not_supported.md): HTTP 400 — No provider supports all required parameters. - [resource_not_found](https://docs.auriko.ai/errors/resource_not_found.md): HTTP 404 — Requested resource not found. - [response_api_only](https://docs.auriko.ai/errors/response_api_only.md): HTTP 400 — Model is only available via the Response API. - [service_unavailable](https://docs.auriko.ai/errors/service_unavailable.md): HTTP 503 — Planned or temporary capacity shortage. - [state_precondition_failed](https://docs.auriko.ai/errors/state_precondition_failed.md): HTTP 409 — Resource is not in the required state for this operation. - [streaming_not_supported](https://docs.auriko.ai/errors/streaming_not_supported.md): HTTP 400 — Model does not support streaming. - [structured_output_not_supported](https://docs.auriko.ai/errors/structured_output_not_supported.md): HTTP 400 — Model does not support structured output. - [thinking_disable_not_supported](https://docs.auriko.ai/errors/thinking_disable_not_supported.md): HTTP 400 — Model does not support disabling thinking. - [throughput_constraint_not_met](https://docs.auriko.ai/errors/throughput_constraint_not_met.md): HTTP 400 — No provider meets the throughput constraint. - [tier_opt_in_required](https://docs.auriko.ai/errors/tier_opt_in_required.md): HTTP 400 — Model requires explicit tier opt-in. - [tool_choice_required_not_supported](https://docs.auriko.ai/errors/tool_choice_required_not_supported.md): HTTP 400 — Model does not support tool_choice="required". - [tools_not_supported](https://docs.auriko.ai/errors/tools_not_supported.md): HTTP 400 — Model does not support tool/function calling. - [tools_with_structured_output_not_supported](https://docs.auriko.ai/errors/tools_with_structured_output_not_supported.md): HTTP 400 — Model does not support tools with structured output. - [unknown_field](https://docs.auriko.ai/errors/unknown_field.md): HTTP 400 — Request contains unrecognized fields. - [unsupported_modalities](https://docs.auriko.ai/errors/unsupported_modalities.md): HTTP 400 — Audio output is not supported. - [upstream_error](https://docs.auriko.ai/errors/upstream_error.md): HTTP 502 — An upstream provider returned an invalid or unparseable response. - [upstream_timeout](https://docs.auriko.ai/errors/upstream_timeout.md): HTTP 504 — Upstream exceeded the request deadline. - [vision_not_supported](https://docs.auriko.ai/errors/vision_not_supported.md): HTTP 400 — Model does not support vision/image input. - [Response Metadata](https://docs.auriko.ai/contract/response-metadata.md): Fields in the routing_metadata object returned with every API response - [Response Headers](https://docs.auriko.ai/contract/response-headers.md): All publicly stable response headers returned by the Auriko API - [Canonical models](https://docs.auriko.ai/contract/canonical-models.md): Query the model directory for a live list of all available canonical model IDs. ## Platform - [Plans and billing](https://docs.auriko.ai/platform/plans.md): Compare Free and Pro plan entitlements, rate limits, and billing. - [Rate limits](https://docs.auriko.ai/platform/rate-limits.md): Check rate limit headers and handle 429 responses. - [Bring your own key](https://docs.auriko.ai/platform/byok.md): Route requests through your own provider API keys. ## Changelog - [Changelog](https://docs.auriko.ai/changelog.md): API, SDK, and model updates ## API Reference - [API Overview](https://docs.auriko.ai/api-reference/overview.md): Overview of the Auriko API, endpoints, and response format. - [Authentication](https://docs.auriko.ai/api-reference/authentication.md): How to authenticate with the Auriko API - [Errors](https://docs.auriko.ai/api-reference/errors.md): Error envelope, codes, and retry behavior - [Create Chat Completion](https://docs.auriko.ai/api-reference/chat-completions.md): Create a model response for a chat conversation with intelligent routing - [Create response](https://docs.auriko.ai/api-reference/create-response.md): Create a model response using the OpenAI Response API format - [List callable models](https://docs.auriko.ai/api-reference/list-callable-models.md): Lists model IDs you can use with authenticated Auriko inference requests. - [Retrieve callable model](https://docs.auriko.ai/api-reference/retrieve-callable-model.md): Returns metadata for one model ID or alias you can use in an inference request. - [Model directory](https://docs.auriko.ai/api-reference/model-directory.md): Lists models with provider routes, context windows, capabilities, modalities, and pricing tiers. - [Canonical model records](https://docs.auriko.ai/api-reference/canonical-model-records.md): Lists normalized model records with canonical IDs, capabilities, and provider availability. - [Provider catalog](https://docs.auriko.ai/api-reference/provider-catalog.md): Lists providers that appear in Auriko model catalog entries. - [Get API key identity](https://docs.auriko.ai/api-reference/get-api-key-identity.md): Discover your workspace, scopes, profile, and rate limits from an API key - [List API keys](https://docs.auriko.ai/api-reference/list-api-keys.md): Returns API keys for a workspace - [Get API key](https://docs.auriko.ai/api-reference/get-api-key.md): Returns a specific API key's metadata - [Get API key usage](https://docs.auriko.ai/api-reference/get-api-key-usage.md): Returns usage statistics for a specific API key - [Create API key](https://docs.auriko.ai/api-reference/create-api-key.md): Creates a new API key in a workspace - [Update API key](https://docs.auriko.ai/api-reference/update-api-key.md): Updates an API key's mutable properties - [Revoke API key](https://docs.auriko.ai/api-reference/delete-api-key.md): Revokes an API key - [List BYOK providers](https://docs.auriko.ai/api-reference/list-byok-providers.md): Lists providers that accept bring-your-own-key credentials - [List provider account tiers](https://docs.auriko.ai/api-reference/get-byok-provider-tiers.md): Lists the allowed account_tier values for a BYOK provider - [List BYOK keys](https://docs.auriko.ai/api-reference/list-byok-keys.md): Lists a workspace's BYOK keys as redacted metadata - [Create BYOK key](https://docs.auriko.ai/api-reference/create-byok-key.md): Creates a BYOK key after validating the provider secret - [Get BYOK key](https://docs.auriko.ai/api-reference/get-byok-key.md): Returns a BYOK key's redacted metadata - [Update BYOK key](https://docs.auriko.ai/api-reference/update-byok-key.md): Updates a BYOK key's metadata and routing state - [Delete BYOK key](https://docs.auriko.ai/api-reference/delete-byok-key.md): Deletes a BYOK key - [Get credit balance](https://docs.auriko.ai/api-reference/get-credit-balance.md): Returns the workspace credit balance and billing configuration ## SDK - [Integration Guide](https://docs.auriko.ai/sdk/integration-guide.md): Choose the right integration channel for your Auriko project - [Python SDK](https://docs.auriko.ai/sdk/python.md): Send requests, stream responses, and route across models with the Python SDK - [Python SDK Reference](https://docs.auriko.ai/sdk/python-reference.md): Look up every type, parameter, and error class in the Python SDK - [TypeScript SDK](https://docs.auriko.ai/sdk/typescript.md): Send requests, stream responses, and route across models with the TypeScript SDK - [TypeScript SDK Reference](https://docs.auriko.ai/sdk/typescript-reference.md): Look up every type, parameter, and error class in the TypeScript SDK ## Full corpus - [Complete documentation](https://docs.auriko.ai/llms-full.txt): Full corpus for single-fetch ingestion (includes OpenAPI spec extraction not present in per-page markdown exports).