# ReqLLM v1.29.0 - Table of Contents

## Pages

- Overview
  - [Overview](overview.md)
  - [ReqLLM 1.x Compatibility Policy](compatibility.md)

- Guides
  - [Getting Started](getting-started-1.md)
  - [Configuration](configuration.md)
  - [Telemetry](telemetry.md)
  - [V1 Telemetry Contract](telemetry-contract.md)
  - [Core Concepts](core-concepts.md)
  - [Data Structures](data-structures.md)
  - [One-call host integration](host-integration.md)
  - [Provider-native integration boundaries](provider-native-integrations.md)
  - [Pricing Policy](pricing-policy.md)
  - [Model Specs](model-specs.md)
  - [Usage & Billing](usage-and-billing.md)
  - [Image Generation](image-generation-2.md)
  - [Model Metadata](model-metadata.md)
  - [Model Support Evidence](model-support.md)
  - [V2 Migration Audit](v2-migration-audit.md)

- Livebooks
  - [ReqLLM • Getting Started](getting-started-3.md)
  - [ReqLLM • Image Generation Comparison](image-generation-4.md)

- Development & Testing
  - [Mix Tasks Guide](mix-tasks.md)
  - [Fixture Testing Guide](fixture-testing.md)
  - [Adding a New Provider](adding_a_provider.md)
  - [Provider Extension Manifest Decision](provider-extension-decision.md)

- Providers
  - [Anthropic](anthropic.md)
  - [OpenAI](openai.md)
  - [Google (Gemini)](google.md)
  - [Azure](azure.md)
  - [Google Vertex AI](google_vertex.md)
  - [xAI (Grok)](xai.md)
  - [Groq](groq.md)
  - [OpenRouter](openrouter.md)
  - [Ollama (Local LLMs)](ollama.md)
  - [LM Studio](lmstudio.md)
  - [Amazon Bedrock](amazon_bedrock.md)
  - [Cerebras](cerebras.md)
  - [DeepSeek](deepseek.md)
  - [StepFun](stepfun.md)
  - [Fireworks AI](fireworks_ai.md)
  - [Meta Model API](meta.md)
  - [GitHub Copilot](github_copilot.md)
  - [Moonshot AI](moonshot_ai.md)
  - [NEAR AI Cloud](nearai.md)
  - [Reflection AI](reflection.md)
  - [Zenmux](zenmux.md)
  - [Z.AI](zai.md)
  - [Z.AI Coder](zai_coder.md)

- Changelog
  - [Changelog](changelog.md)

- Contributing
  - [Contributing to ReqLLM](contributing.md)

## Modules

- [ReqLLM.Bedrock.BidiStream](ReqLLM.Bedrock.BidiStream.md): Low-level client for AWS Bedrock `InvokeModelWithBidirectionalStream`
(e.g. Amazon Nova Sonic speech-to-speech).
- [ReqLLM.Bedrock.BidiStream.Conn](ReqLLM.Bedrock.BidiStream.Conn.md): Handle to a running bidirectional Bedrock session.
- [ReqLLM.Bedrock.NovaSonic](ReqLLM.Bedrock.NovaSonic.md): Amazon Nova Sonic speech-to-speech client over Bedrock bidirectional streaming.
- [ReqLLM.Bedrock.NovaSonic.Session](ReqLLM.Bedrock.NovaSonic.Session.md): A running Nova Sonic conversation.
- [ReqLLM.Cache](ReqLLM.Cache.md): Application-layer response cache hooks for generation requests.
- [ReqLLM.Compaction](ReqLLM.Compaction.md): Context compaction for the OpenAI Responses API (OpenAI and Azure OpenAI).
- [ReqLLM.Compatibility.Evidence](ReqLLM.Compatibility.Evidence.md): Versioned compatibility evidence for models, execution surfaces, and scenarios.
- [ReqLLM.Compatibility.ProviderDrift](ReqLLM.Compatibility.ProviderDrift.md): Bounded, read-only live verification for a small compatibility anchor matrix.
- [ReqLLM.Compatibility.ScenarioCatalog](ReqLLM.Compatibility.ScenarioCatalog.md): Compiled catalog of compatibility scenarios used by ReqLLM's test tooling.
- [ReqLLM.Compatibility.SupportReference](ReqLLM.Compatibility.SupportReference.md): Deterministic Markdown generation from compatibility catalog and evidence.

- [ReqLLM.Doctor](ReqLLM.Doctor.md): Read-only installation and runtime diagnostics for ReqLLM.
- [ReqLLM.Evaluation](ReqLLM.Evaluation.md): Evaluates one text or JSON-compatible state against named questions.
- [ReqLLM.FinchRequestAdapter](ReqLLM.FinchRequestAdapter.md): Behaviour for transforming `Finch.Request` structs just before a streaming
request is sent.
- [ReqLLM.Images.OpenAICompatible](ReqLLM.Images.OpenAICompatible.md): Shared codec for providers that speak the OpenAI Images wire format.
- [ReqLLM.OCR](ReqLLM.OCR.md): Optical Character Recognition for ReqLLM.
- [ReqLLM.OpenAI.Realtime](ReqLLM.OpenAI.Realtime.md): Experimental low-level Realtime WebSocket client for OpenAI.
- [ReqLLM.OpenAI.Realtime.Event](ReqLLM.OpenAI.Realtime.Event.md): An experimental view of one OpenAI Realtime server event.
- [ReqLLM.OpenAI.Realtime.Session](ReqLLM.OpenAI.Realtime.Session.md): An experimental OpenAI Realtime WebSocket session.

- [ReqLLM.OpenAI.Responses](ReqLLM.OpenAI.Responses.md): Experimental OpenAI Responses helpers and persistent WebSocket sessions.
- [ReqLLM.OpenAI.Responses.Session](ReqLLM.OpenAI.Responses.Session.md): An experimental OpenAI Responses WebSocket session.
- [ReqLLM.OpenTelemetry.Attributes](ReqLLM.OpenTelemetry.Attributes.md): Builds the scalar `gen_ai.*` / `server.*` / `error.*` attribute maps
emitted on GenAI client spans, from ReqLLM request lifecycle metadata.
- [ReqLLM.OpenTelemetry.Content](ReqLLM.OpenTelemetry.Content.md): Shapes ReqLLM request and response payloads into the GenAI content
attributes — `gen_ai.input.messages`, `gen_ai.system_instructions`,
`gen_ai.tool.definitions`, `gen_ai.output.messages`.
- [ReqLLM.OpenTelemetry.Metrics](ReqLLM.OpenTelemetry.Metrics.md): Builds histogram records for the four OpenTelemetry GenAI client metrics:
`gen_ai.client.operation.duration`, `gen_ai.client.token.usage`,
`gen_ai.client.operation.time_to_first_chunk`, and
`gen_ai.client.operation.time_per_output_chunk`.
- [ReqLLM.OpenTelemetry.SemConv](ReqLLM.OpenTelemetry.SemConv.md): Spec name tables for the OpenTelemetry GenAI semantic conventions —
`gen_ai.provider.name`, `gen_ai.operation.name`, `gen_ai.output.type`,
and the canonical span name.
- [ReqLLM.OpenTelemetry.Shared](ReqLLM.OpenTelemetry.Shared.md): Cross-cutting helpers shared by `ReqLLM.OpenTelemetry` and
`ReqLLM.Telemetry.OpenTelemetry` — option parsing, Langfuse cost-details
merging, error rendering.
- [ReqLLM.Output](ReqLLM.Output.md): Describes the value expected from text generation.
- [ReqLLM.Output.Result](ReqLLM.Output.Result.md): Describes how a complete `ReqLLM.Output` value was materialized and validated.
- [ReqLLM.PricingContext](ReqLLM.PricingContext.md): Confirmed request and response facts used for local pricing.
- [ReqLLM.Provider.ChunkAccumulator](ReqLLM.Provider.ChunkAccumulator.md): Shared streaming-chunk reducer used by both `ReqLLM.StreamServer` (the
hot path, one chunk at a time) and
`ReqLLM.Provider.Defaults.ResponseBuilder` (batch, full chunk list at
end-of-stream).
- [ReqLLM.Provider.InProcessStream](ReqLLM.Provider.InProcessStream.md): Describes a provider-owned in-process stream of canonical chunks.
- [ReqLLM.ProviderFileReference](ReqLLM.ProviderFileReference.md): Metadata and validation for explicitly provider-owned file references.
- [ReqLLM.Rerank](ReqLLM.Rerank.md): Reranking functionality for ReqLLM.
- [ReqLLM.RerankResponse](ReqLLM.RerankResponse.md): Canonical reranking response for ReqLLM.
- [ReqLLM.Response.OutputItem](ReqLLM.Response.OutputItem.md): A computed, canonical view of one value produced by a model call.
- [ReqLLM.Router](ReqLLM.Router.md): A behaviour for application-defined model routers.
- [ReqLLM.Router.Request](ReqLLM.Router.Request.md): The normalized input for an application model router.
- [ReqLLM.Schema.EvictionStrategy](ReqLLM.Schema.EvictionStrategy.md): Selects an entry to remove from the JSON Schema validator cache.
- [ReqLLM.Schema.EvictionStrategy.FIFO](ReqLLM.Schema.EvictionStrategy.FIFO.md): Removes the JSON Schema validator that entered the cache first.

- [ReqLLM.Schema.EvictionStrategy.LRU](ReqLLM.Schema.EvictionStrategy.LRU.md): Removes the least recently used JSON Schema validator.

- [ReqLLM.StreamEvent](ReqLLM.StreamEvent.md): A tagged, provider-neutral projection of one streaming event.
- [ReqLLM.Telemetry.RequestOptions](ReqLLM.Telemetry.RequestOptions.md): Normalizes caller-facing inference options into the compact `request_options`
map exposed on `[:req_llm, :request, *]` telemetry metadata.
- [ReqLLM.Telemetry.RequestSource](ReqLLM.Telemetry.RequestSource.md): Extracts the upstream server (`address`, `port`, `path`) from a request
representation for the telemetry context's `server` map.
- [ReqLLM.ToolCallIdCompat](ReqLLM.ToolCallIdCompat.md): Tool call ID compatibility helpers for cross-provider conversations.
- [ReqLLM.Video](ReqLLM.Video.md): Video generation functionality for ReqLLM.
- [ReqLLM.Video.File](ReqLLM.Video.File.md): A resolved video file.
- [ReqLLM.Video.Task](ReqLLM.Video.Task.md): A video generation task.

- Top-Level API
  - [ReqLLM](ReqLLM.md): Main API facade for Req AI.
  - [ReqLLM.Context](ReqLLM.Context.md): Context represents a conversation history as a collection of messages.
  - [ReqLLM.Images](ReqLLM.Images.md): Image generation functionality for ReqLLM.
  - [ReqLLM.Schema](ReqLLM.Schema.md): Single schema authority for NimbleOptions ↔ JSON Schema conversion.

- Utilities
  - [ReqLLM.Debug](ReqLLM.Debug.md): Centralized debug logging for ReqLLM development and troubleshooting.
  - [ReqLLM.Error](ReqLLM.Error.md): Error handling system for ReqLLM using Splode.

  - [ReqLLM.Keys](ReqLLM.Keys.md): Handles API key lookup with the following precedence
  - [ReqLLM.Migration](ReqLLM.Migration.md): Read-only V2 migration inventory for ReqLLM applications.
  - [ReqLLM.ModelHelpers](ReqLLM.ModelHelpers.md): Helper functions for querying LLMDB.Model capabilities.
  - [ReqLLM.OpenTelemetry](ReqLLM.OpenTelemetry.md): Bridges ReqLLM request lifecycle telemetry into OpenTelemetry GenAI client spans.
  - [ReqLLM.OpenTelemetry.Adapter](ReqLLM.OpenTelemetry.Adapter.md): Behaviour the OpenTelemetry bridge uses to talk to a tracer.
  - [ReqLLM.OpenTelemetry.OTelAdapter](ReqLLM.OpenTelemetry.OTelAdapter.md): Default `ReqLLM.OpenTelemetry.Adapter` implementation, backed by the
Erlang OpenTelemetry SDK (`:otel_tracer`, `:otel_span`, `:otel_meter`).
  - [ReqLLM.ParamTransform](ReqLLM.ParamTransform.md): Composable parameter transformation engine for applying model-specific rules to options.
  - [ReqLLM.Telemetry](ReqLLM.Telemetry.md): Native `:telemetry` emitter for ReqLLM request and reasoning lifecycle.
  - [ReqLLM.Telemetry.OpenTelemetry](ReqLLM.Telemetry.OpenTelemetry.md): Dependency-free helpers for mapping ReqLLM telemetry metadata to
OpenTelemetry GenAI span data.
  - [ReqLLM.Usage](ReqLLM.Usage.md): Usage normalization helpers.

- Data Structures
  - [ReqLLM.Embedding](ReqLLM.Embedding.md): Embedding functionality for ReqLLM.
  - [ReqLLM.Generation](ReqLLM.Generation.md): Text generation functionality for ReqLLM.
  - [ReqLLM.Message](ReqLLM.Message.md): Message represents a single conversation message with multi-modal content support.
  - [ReqLLM.Message.ContentPart](ReqLLM.Message.ContentPart.md): ContentPart represents a single piece of content within a message.
  - [ReqLLM.Message.ReasoningDetails](ReqLLM.Message.ReasoningDetails.md): Normalized reasoning/thinking data from LLM providers.
  - [ReqLLM.Response](ReqLLM.Response.md): High-level representation of an LLM response.
  - [ReqLLM.Response.Stream](ReqLLM.Response.Stream.md): Stream processing utilities for ReqLLM responses.
  - [ReqLLM.StreamChunk](ReqLLM.StreamChunk.md): Represents a single chunk in a streaming response.
  - [ReqLLM.StreamResponse](ReqLLM.StreamResponse.md): A streaming response container that provides both real-time streaming and asynchronous metadata.
  - [ReqLLM.StreamResponse.MetadataHandle](ReqLLM.StreamResponse.MetadataHandle.md): Asynchronous metadata cache that allows multiple awaiters to share the same result.
  - [ReqLLM.Tool](ReqLLM.Tool.md): Tool definition for AI model function calling.
  - [ReqLLM.ToolCall](ReqLLM.ToolCall.md): Represents a single tool call from an assistant message.
  - [ReqLLM.ToolResult](ReqLLM.ToolResult.md): ToolResult represents structured and multi-part tool outputs while keeping
application-visible data separate from model-facing content.

- Steps
  - [ReqLLM.Step.Error](ReqLLM.Step.Error.md): Req step that integrates with Splode error handling.
  - [ReqLLM.Step.Fixture](ReqLLM.Step.Fixture.md): Req step that attaches test fixture functionality when running in test environments.
  - [ReqLLM.Step.Retry](ReqLLM.Step.Retry.md): Req step that handles automatic retries for transient failures.
  - [ReqLLM.Step.Telemetry](ReqLLM.Step.Telemetry.md): Req step that emits request and reasoning telemetry for Req-backed flows.

  - [ReqLLM.Step.Usage](ReqLLM.Step.Usage.md): Centralized Req step that extracts token usage from provider responses,
normalizes usage values across providers, computes costs, and emits telemetry.

- Streaming
  - [ReqLLM.StreamServer](ReqLLM.StreamServer.md): GenServer that manages streaming LLM sessions with backpressure and SSE parsing.
  - [ReqLLM.Streaming](ReqLLM.Streaming.md): Main orchestration for ReqLLM streaming operations.
  - [ReqLLM.Streaming.FinchClient](ReqLLM.Streaming.FinchClient.md): Finch HTTP client for ReqLLM streaming operations.
  - [ReqLLM.Streaming.Fixtures.HTTPContext](ReqLLM.Streaming.Fixtures.HTTPContext.md): Lightweight HTTP context for streaming operations.
  - [ReqLLM.Streaming.Retry](ReqLLM.Streaming.Retry.md): Retry wrapper for Finch streaming requests.
  - [ReqLLM.Streaming.SSE](ReqLLM.Streaming.SSE.md): Provider-agnostic Server-Sent Events (SSE) parsing utilities.

- Transcription
  - [ReqLLM.Transcription](ReqLLM.Transcription.md): Speech-to-text transcription functionality for ReqLLM.
  - [ReqLLM.Transcription.DetailedResult](ReqLLM.Transcription.DetailedResult.md): Opt-in transcription result with sparse call metadata.
  - [ReqLLM.Transcription.Result](ReqLLM.Transcription.Result.md): Result of an audio transcription operation.

- Speech
  - [ReqLLM.Speech](ReqLLM.Speech.md): Text-to-speech generation functionality for ReqLLM.
  - [ReqLLM.Speech.DetailedResult](ReqLLM.Speech.DetailedResult.md): Opt-in speech result with sparse call metadata.
  - [ReqLLM.Speech.Result](ReqLLM.Speech.Result.md): Result of a text-to-speech generation operation.

- Provider Extension API
  - [ReqLLM.Provider](ReqLLM.Provider.md): Behavior for LLM provider implementations.
  - [ReqLLM.Provider.Defaults](ReqLLM.Provider.Defaults.md): Default implementations for common provider behavior patterns.
  - [ReqLLM.Provider.Defaults.ResponseBuilder](ReqLLM.Provider.Defaults.ResponseBuilder.md): Default ResponseBuilder implementation for OpenAI-compatible providers.
  - [ReqLLM.Provider.Options](ReqLLM.Provider.Options.md): Runtime generation options processing for ReqLLM providers.
  - [ReqLLM.Provider.ResponseBuilder](ReqLLM.Provider.ResponseBuilder.md): Behaviour for provider-specific Response assembly from StreamChunks.
  - [ReqLLM.Provider.Utils](ReqLLM.Provider.Utils.md): Shared utilities for provider implementations.
  - [ReqLLM.Providers](ReqLLM.Providers.md): Provider discovery and dispatch via introspection.

- Providers
  - [ReqLLM.Providers.Alibaba](ReqLLM.Providers.Alibaba.md): Alibaba Cloud Bailian (DashScope) provider – international endpoint.
  - [ReqLLM.Providers.Alibaba.Shared](ReqLLM.Providers.Alibaba.Shared.md): Shared logic for Alibaba Cloud Bailian (DashScope) providers.
  - [ReqLLM.Providers.AlibabaCN](ReqLLM.Providers.AlibabaCN.md): Alibaba Cloud Bailian (DashScope) provider – China/Beijing endpoint.
  - [ReqLLM.Providers.AmazonBedrock](ReqLLM.Providers.AmazonBedrock.md): AWS Bedrock provider implementation using the Provider behavior.
  - [ReqLLM.Providers.AmazonBedrock.AWSEventStream](ReqLLM.Providers.AmazonBedrock.AWSEventStream.md): Parser for the AWS Event Stream protocol, specialized for Amazon Bedrock.
  - [ReqLLM.Providers.AmazonBedrock.Anthropic](ReqLLM.Providers.AmazonBedrock.Anthropic.md): Anthropic model family support for AWS Bedrock.
  - [ReqLLM.Providers.AmazonBedrock.Cohere](ReqLLM.Providers.AmazonBedrock.Cohere.md): Cohere model family support for AWS Bedrock.
  - [ReqLLM.Providers.AmazonBedrock.Converse](ReqLLM.Providers.AmazonBedrock.Converse.md): AWS Bedrock Converse API support for unified tool calling across models.
  - [ReqLLM.Providers.AmazonBedrock.Meta](ReqLLM.Providers.AmazonBedrock.Meta.md): Meta Llama model family support for AWS Bedrock.
  - [ReqLLM.Providers.AmazonBedrock.OpenAI](ReqLLM.Providers.AmazonBedrock.OpenAI.md): OpenAI model family support for AWS Bedrock.
  - [ReqLLM.Providers.AmazonBedrock.PromptCache](ReqLLM.Providers.AmazonBedrock.PromptCache.md): Prompt caching options shared by the Amazon Bedrock formatters.
  - [ReqLLM.Providers.AmazonBedrock.Response](ReqLLM.Providers.AmazonBedrock.Response.md): Shared utilities for unwrapping AWS Bedrock response formats.
  - [ReqLLM.Providers.AmazonBedrock.Responses](ReqLLM.Providers.AmazonBedrock.Responses.md): Bedrock Mantle adapter for the OpenAI Responses API wire format.
  - [ReqLLM.Providers.AmazonBedrock.STS](ReqLLM.Providers.AmazonBedrock.STS.md): AWS Security Token Service (STS) integration for AssumeRole.
  - [ReqLLM.Providers.Anthropic](ReqLLM.Providers.Anthropic.md): Provider implementation for Anthropic Claude models.
  - [ReqLLM.Providers.Anthropic.AdapterHelpers](ReqLLM.Providers.Anthropic.AdapterHelpers.md): Shared helper functions for Anthropic model adapters (Bedrock, Vertex).
  - [ReqLLM.Providers.Anthropic.Context](ReqLLM.Providers.Anthropic.Context.md): Anthropic-specific context encoding for the Messages API format.
  - [ReqLLM.Providers.Anthropic.PlatformReasoning](ReqLLM.Providers.Anthropic.PlatformReasoning.md): Shared extended thinking/reasoning support for Anthropic models on third-party platforms.
  - [ReqLLM.Providers.Anthropic.Response](ReqLLM.Providers.Anthropic.Response.md): Anthropic-specific response decoding for the Messages API format.
  - [ReqLLM.Providers.Anthropic.ResponseBuilder](ReqLLM.Providers.Anthropic.ResponseBuilder.md): Anthropic-specific ResponseBuilder implementation.
  - [ReqLLM.Providers.AtlasCloud](ReqLLM.Providers.AtlasCloud.md): Atlas Cloud provider using the OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.Azure](ReqLLM.Providers.Azure.md): Azure AI provider implementation.
  - [ReqLLM.Providers.Azure.Anthropic](ReqLLM.Providers.Azure.Anthropic.md): Anthropic model family support for Azure.
  - [ReqLLM.Providers.Azure.OpenAI](ReqLLM.Providers.Azure.OpenAI.md): OpenAI model family support for Azure OpenAI Service.
  - [ReqLLM.Providers.Azure.ResponsesAPI](ReqLLM.Providers.Azure.ResponsesAPI.md): Azure Responses API adapter.
  - [ReqLLM.Providers.CatalogGateway](ReqLLM.Providers.CatalogGateway.md): Runs OpenAI Chat Completions models for providers with LLMDB execution metadata.
  - [ReqLLM.Providers.Cerebras](ReqLLM.Providers.Cerebras.md): Cerebras provider – OpenAI-compatible Chat Completions API with ultra-fast inference.
  - [ReqLLM.Providers.Cohere](ReqLLM.Providers.Cohere.md): Cohere provider implementation for reranking operations.

  - [ReqLLM.Providers.Deepseek](ReqLLM.Providers.Deepseek.md): DeepSeek AI provider – OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.ElevenLabs](ReqLLM.Providers.ElevenLabs.md): ElevenLabs provider for text-to-speech and speech-to-text transcription.
  - [ReqLLM.Providers.FireworksAI](ReqLLM.Providers.FireworksAI.md): Fireworks AI provider – OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.GitHubCopilot](ReqLLM.Providers.GitHubCopilot.md): GitHub Copilot provider using Copilot's OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.Google](ReqLLM.Providers.Google.md): Google Gemini provider – built on the OpenAI baseline defaults with Gemini-specific customizations.
  - [ReqLLM.Providers.Google.CachedContent](ReqLLM.Providers.Google.CachedContent.md): Shared functionality for Google's Context Caching API.
  - [ReqLLM.Providers.Google.ResponseBuilder](ReqLLM.Providers.Google.ResponseBuilder.md): Google/Gemini-specific ResponseBuilder implementation.
  - [ReqLLM.Providers.GoogleVertex](ReqLLM.Providers.GoogleVertex.md): Google Vertex AI provider implementation.
  - [ReqLLM.Providers.GoogleVertex.Anthropic](ReqLLM.Providers.GoogleVertex.Anthropic.md): Anthropic model family support for Google Vertex AI.
  - [ReqLLM.Providers.GoogleVertex.Auth](ReqLLM.Providers.GoogleVertex.Auth.md): Google Cloud OAuth2 authentication for Vertex AI.
  - [ReqLLM.Providers.GoogleVertex.Gemini](ReqLLM.Providers.GoogleVertex.Gemini.md): Gemini model family support for Google Vertex AI.
  - [ReqLLM.Providers.GoogleVertex.OpenAICompat](ReqLLM.Providers.GoogleVertex.OpenAICompat.md): OpenAI-compatible model family support for Google Vertex AI.
  - [ReqLLM.Providers.GoogleVertex.TokenCache](ReqLLM.Providers.GoogleVertex.TokenCache.md): OAuth2 token cache for Google Vertex AI.
  - [ReqLLM.Providers.Groq](ReqLLM.Providers.Groq.md): Groq provider – 100% OpenAI Chat Completions compatible with Groq's high-performance hardware.
  - [ReqLLM.Providers.LMStudio](ReqLLM.Providers.LMStudio.md): Local inference through LM Studio's OpenAI-compatible API.
  - [ReqLLM.Providers.Meta](ReqLLM.Providers.Meta.md): Meta Model API provider using the OpenAI-compatible Responses API.
  - [ReqLLM.Providers.Meta.Llama](ReqLLM.Providers.Meta.Llama.md): Meta Llama native-payload helpers used by Amazon Bedrock.
  - [ReqLLM.Providers.Minimax](ReqLLM.Providers.Minimax.md): MiniMax provider using the OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.Minimax.ImagesAPI](ReqLLM.Providers.Minimax.ImagesAPI.md): MiniMax Images API driver.
  - [ReqLLM.Providers.Minimax.ResponseBuilder](ReqLLM.Providers.Minimax.ResponseBuilder.md): MiniMax-specific ResponseBuilder implementation.
  - [ReqLLM.Providers.Minimax.VideoAPI](ReqLLM.Providers.Minimax.VideoAPI.md): MiniMax Video API driver.
  - [ReqLLM.Providers.Minimax.VideoAPIV1](ReqLLM.Providers.Minimax.VideoAPIV1.md): MiniMax Video API V1 driver (Hailuo series).
  - [ReqLLM.Providers.Mistral](ReqLLM.Providers.Mistral.md): Mistral provider built on the official Mistral chat completions and embeddings APIs.
  - [ReqLLM.Providers.MoonshotAI](ReqLLM.Providers.MoonshotAI.md): Moonshot AI provider using the OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.NearAI](ReqLLM.Providers.NearAI.md): NEAR AI Cloud provider using the OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.Ollama](ReqLLM.Providers.Ollama.md): Ollama provider — local LLM inference via Ollama's OpenAI-compatible API.
  - [ReqLLM.Providers.OpenAI](ReqLLM.Providers.OpenAI.md): OpenAI provider implementation with multi-driver architecture for Chat, Responses, Decisions,
and Images APIs.
  - [ReqLLM.Providers.OpenAI.API](ReqLLM.Providers.OpenAI.API.md): Behaviour for OpenAI API endpoint drivers.
  - [ReqLLM.Providers.OpenAI.AdapterHelpers](ReqLLM.Providers.OpenAI.AdapterHelpers.md): Shared helper functions for OpenAI-compatible model adapters (Azure, etc.).
  - [ReqLLM.Providers.OpenAI.ChatAPI](ReqLLM.Providers.OpenAI.ChatAPI.md): OpenAI Chat Completions API driver.
  - [ReqLLM.Providers.OpenAI.DecisionsAPI](ReqLLM.Providers.OpenAI.DecisionsAPI.md): OpenAI Decisions endpoint driver.
  - [ReqLLM.Providers.OpenAI.Files](ReqLLM.Providers.OpenAI.Files.md): OpenAI-scoped lifecycle operations for reusable provider files.
  - [ReqLLM.Providers.OpenAI.Files.Page](ReqLLM.Providers.OpenAI.Files.Page.md): A page returned by `ReqLLM.Providers.OpenAI.Files.list/1`.
  - [ReqLLM.Providers.OpenAI.ImagesAPI](ReqLLM.Providers.OpenAI.ImagesAPI.md): OpenAI Images API driver.
  - [ReqLLM.Providers.OpenAI.ParamProfiles](ReqLLM.Providers.OpenAI.ParamProfiles.md): Defines reusable parameter transformation profiles for OpenAI models.
  - [ReqLLM.Providers.OpenAI.ResponsesAPI](ReqLLM.Providers.OpenAI.ResponsesAPI.md): OpenAI Responses API driver for Responses endpoint models.
  - [ReqLLM.Providers.OpenAI.ResponsesAPI.ResponseBuilder](ReqLLM.Providers.OpenAI.ResponsesAPI.ResponseBuilder.md): OpenAI Responses API-specific ResponseBuilder implementation.
  - [ReqLLM.Providers.OpenAICodex](ReqLLM.Providers.OpenAICodex.md): OpenAI Codex provider backed by the ChatGPT Codex responses endpoint.
  - [ReqLLM.Providers.OpenRouter](ReqLLM.Providers.OpenRouter.md): OpenRouter provider – OpenAI Chat Completions compatible with OpenRouter's unified API.
  - [ReqLLM.Providers.Reflection](ReqLLM.Providers.Reflection.md): Reflection AI provider for text, tool calls, and structured output.
  - [ReqLLM.Providers.StepFun](ReqLLM.Providers.StepFun.md): StepFun China provider for chat, speech generation, and transcription.
  - [ReqLLM.Providers.StepFunAI](ReqLLM.Providers.StepFunAI.md): StepFun Global provider for chat, speech generation, and transcription.
  - [ReqLLM.Providers.TypeSafe](ReqLLM.Providers.TypeSafe.md): TypeSafe evaluation provider for Jev and later System One models.
  - [ReqLLM.Providers.VLLM](ReqLLM.Providers.VLLM.md): vLLM provider – self-hosted OpenAI-compatible Chat Completions API.
  - [ReqLLM.Providers.Venice](ReqLLM.Providers.Venice.md): Venice AI provider – OpenAI-compatible Chat Completions API with privacy-first inference.
  - [ReqLLM.Providers.XAI](ReqLLM.Providers.XAI.md): xAI (Grok) provider – OpenAI Chat Completions compatible with xAI's models and features.
  - [ReqLLM.Providers.XAI.ImagesAPI](ReqLLM.Providers.XAI.ImagesAPI.md): xAI Images API driver.
  - [ReqLLM.Providers.Zai](ReqLLM.Providers.Zai.md): Z.AI provider – OpenAI-compatible Chat Completions API (Standard Endpoint).
  - [ReqLLM.Providers.ZaiCoder](ReqLLM.Providers.ZaiCoder.md): Z.AI Coder provider – OpenAI-compatible Chat Completions API (Coding Endpoint).
  - [ReqLLM.Providers.ZaiCodingPlan](ReqLLM.Providers.ZaiCodingPlan.md): Z.AI Coding Plan provider – alias for zai_coder.
  - [ReqLLM.Providers.Zenmux](ReqLLM.Providers.Zenmux.md): Zenmux provider – OpenAI Chat Completions compatible with Zenmux's unified API.

- Exceptions
  - [ReqLLM.Error.API](ReqLLM.Error.API.md): Error class for API-related failures and HTTP errors.
  - [ReqLLM.Error.API.JSONDecode](ReqLLM.Error.API.JSONDecode.md): Error for when we can't parse the JSON response.
  - [ReqLLM.Error.API.Request](ReqLLM.Error.API.Request.md): Error for API request failures, HTTP errors, and network issues.
  - [ReqLLM.Error.API.Response](ReqLLM.Error.API.Response.md): Error for provider response parsing failures and unexpected response formats.
  - [ReqLLM.Error.API.SchemaValidation](ReqLLM.Error.API.SchemaValidation.md): Error for when generated objects don't match the expected schema.
  - [ReqLLM.Error.API.Stream](ReqLLM.Error.API.Stream.md): Error for stream processing failures.
  - [ReqLLM.Error.API.Timeout](ReqLLM.Error.API.Timeout.md): Structured timeout for ReqLLM model calls and provider transports.
  - [ReqLLM.Error.Invalid](ReqLLM.Error.Invalid.md): Error class for invalid input parameters and configurations.
  - [ReqLLM.Error.Invalid.Capability](ReqLLM.Error.Invalid.Capability.md): Error for unsupported model capabilities.
  - [ReqLLM.Error.Invalid.Content](ReqLLM.Error.Invalid.Content.md): Error for invalid message content.
  - [ReqLLM.Error.Invalid.Message](ReqLLM.Error.Invalid.Message.md): Error for invalid message structures or validation failures.
  - [ReqLLM.Error.Invalid.MessageList](ReqLLM.Error.Invalid.MessageList.md): Error for invalid message list structures.
  - [ReqLLM.Error.Invalid.NotImplemented](ReqLLM.Error.Invalid.NotImplemented.md): Error for unimplemented functionality.
  - [ReqLLM.Error.Invalid.Parameter](ReqLLM.Error.Invalid.Parameter.md): Error for invalid or missing parameters.
  - [ReqLLM.Error.Invalid.Provider](ReqLLM.Error.Invalid.Provider.md): Error for unknown or unsupported providers.
  - [ReqLLM.Error.Invalid.Provider.NotImplemented](ReqLLM.Error.Invalid.Provider.NotImplemented.md): Error for providers that exist but have no implementation (metadata-only).
  - [ReqLLM.Error.Invalid.ProviderFileReference](ReqLLM.Error.Invalid.ProviderFileReference.md): Error for an explicitly owned provider file used outside its lifecycle.
  - [ReqLLM.Error.Invalid.Role](ReqLLM.Error.Invalid.Role.md): Error for invalid message roles.
  - [ReqLLM.Error.Invalid.Schema](ReqLLM.Error.Invalid.Schema.md): Error for invalid schema definitions.
  - [ReqLLM.Error.Unknown](ReqLLM.Error.Unknown.md): Error class for unexpected or unhandled errors.
  - [ReqLLM.Error.Unknown.Unknown](ReqLLM.Error.Unknown.Unknown.md): Error for unexpected or unhandled errors.
  - [ReqLLM.Error.Validation](ReqLLM.Error.Validation.md): Error class for validation failures and parameter errors.
  - [ReqLLM.Error.Validation.Error](ReqLLM.Error.Validation.Error.md): Error for parameter validation failures.

## Mix Tasks

- [mix req_llm.billing](Mix.Tasks.ReqLlm.Billing.md): Run the ExUnit billing suite or inspect its saved provider evidence.
- [mix req_llm.doctor](Mix.Tasks.ReqLlm.Doctor.md): Runs read-only ReqLLM installation and runtime diagnostics.
- [mix req_llm.gen](Mix.Tasks.ReqLlm.Gen.md): Generate text or structured objects from any supported AI model with unified interface.
- [mix req_llm.install](Mix.Tasks.ReqLlm.Install.md): Install and configure ReqLLM for use in an application.
- [mix req_llm.migration_audit](Mix.Tasks.ReqLlm.MigrationAudit.md): Scans Elixir source for precise ReqLLM V2 migration patterns without evaluating
or rewriting application code.
- [mix req_llm.model_compat](Mix.Tasks.ReqLlm.ModelCompat.md): Validate ReqLLM model coverage using the fixture system.
- [mix req_llm.model_support](Mix.Tasks.ReqLlm.ModelSupport.md): Inspect evidence-backed support tiers without changing model resolution.
- [mix req_llm.provider_drift](Mix.Tasks.ReqLlm.ProviderDrift.md): Run a small, credential-aware live verification matrix without updating
fixtures or checked-in compatibility evidence.
- [mix test.livebooks](Mix.Tasks.Test.Livebooks.md): Validates livebook files by extracting Elixir code blocks.