Skip to content
J·BJon BishopFractional growth marketer

Artificial intelligence · llms.txt file breakdown

Venice llms.txt

A directly retrieved Markdown index for Venice's documentation, with 147 links across 5 detected headings.

The actual file

Source Markdown

The saved llms.txt text, displayed as source. The file belongs to its publisher; the live endpoint may change.

Open saved text ↗

Showing the first 18000 characters. Load the complete file to read the remainder.

# Venice API

> Venice is a privacy-first, uncensored AI API platform offering text generation, image generation, audio synthesis, speech-to-speech voice conversion, video generation, music, embeddings, and developer tools (web search, document parsing, blockchain RPC) with zero data retention and OpenAI SDK compatibility.

Venice provides permissionless access to AI models with no content filtering, making it ideal for developers building applications that require uncensored outputs, privacy guarantees, and full control over AI interactions. The API is fully compatible with OpenAI's SDK—just change the base URL to `https://api.venice.ai/api/v1`. Venice also offers developer tools including web search, web scraping, document parsing, and blockchain RPC—see the **Tools** section below.

Venice offers four tiers of privacy: **Anonymized** (third-party models with identifying metadata stripped), **Private** (zero data retention, self-hosted open-source models), **TEE** (models running inside hardware-secured enclaves—Venice cannot access the computation), and **E2EE** (end-to-end encrypted models where prompts are encrypted client-side before being sent, and only the TEE can decrypt them).

**AI agents:** see [agents.md](https://docs.venice.ai/agents.md) for a token-optimized agent guide, [claude.md](https://docs.venice.ai/claude.md) for Claude-specific notes, and [skill.md](https://docs.venice.ai/skill.md) for an installable Agent Skill (`npx skills add docs.venice.ai`). Any docs page is available as raw Markdown by appending `.md` to its URL.

## Getting Started

- [Quick Start Guide](https://docs.venice.ai/getting-started/quick-start): Get your API key and make your first request in minutes
- [API Key Generation](https://docs.venice.ai/guides/getting-started/generating-api-key): Step-by-step guide to creating API keys
- [About Venice](https://docs.venice.ai/overview/about-venice): Overview of Venice's capabilities and OpenAI compatibility
- [OpenAI Migration](https://docs.venice.ai/guides/getting-started/openai-migration): Migrate from OpenAI to Venice
- [Postman Collection](https://docs.venice.ai/guides/getting-started/postman): Import ready-to-use API examples

## API Reference

### Text & Chat
- [API Specification](https://docs.venice.ai/api-reference/api-spec): Complete API specification with Venice-specific parameters
- [Chat Completions](https://docs.venice.ai/api-reference/endpoint/chat/completions): Text generation endpoint with streaming, vision, audio input, video input, and tool calling
- [Responses API (Beta)](https://docs.venice.ai/guides/features/responses-api): OpenAI-compatible `POST /responses` endpoint with typed output items for reasoning, messages, function calls, and web search. Stateless, supports streaming via SSE, API key or x402 wallet auth. OpenAI models are served natively (instructions, structured outputs, custom and Codex tools, encrypted reasoning replay). Other models are translated through Chat Completions, where `text.format` and hosted tools are ignored; see the limitations table
- [Model Feature Suffixes](https://docs.venice.ai/api-reference/endpoint/chat/model_feature_suffix): Enable features via model name suffixes as `key=value` pairs (e.g., `model-name:enable_web_search=on` for web search)

### Image
- [Image Generation](https://docs.venice.ai/api-reference/endpoint/image/generate): Create images from text prompts
- [Image Generations (OpenAI-compatible)](https://docs.venice.ai/api-reference/endpoint/image/generations): OpenAI-compatible image generation endpoint
- [Image Upscaling](https://docs.venice.ai/api-reference/endpoint/image/upscale): Enhance and upscale images
- [Image Editing](https://docs.venice.ai/api-reference/endpoint/image/edit): AI-powered image inpainting
- [Image Multi-Edit](https://docs.venice.ai/api-reference/endpoint/image/multi-edit): Combine and edit multiple images with layered inputs (limit is per model; see `model_spec.constraints.maxInputImages`)
- [Background Remove](https://docs.venice.ai/api-reference/endpoint/image/background-remove): Remove image backgrounds
- [Image Styles](https://docs.venice.ai/api-reference/endpoint/image/styles): List available image generation styles

### Audio
- [Text-to-Speech](https://docs.venice.ai/api-reference/endpoint/audio/speech): Convert text to audio with 50+ voices
- [Speech-to-Text](https://docs.venice.ai/api-reference/endpoint/audio/transcriptions): Transcribe audio to text with optional word-level timestamps
- [Music Generation (Queue)](https://docs.venice.ai/api-reference/endpoint/audio/queue): Queue music/sound effects generation
- [Music Generation (Retrieve)](https://docs.venice.ai/api-reference/endpoint/audio/retrieve): Retrieve generated audio by ID
- [Music Generation (Quote)](https://docs.venice.ai/api-reference/endpoint/audio/quote): Get price quote for audio generation
- [Music Generation (Complete)](https://docs.venice.ai/api-reference/endpoint/audio/complete): Queue and wait for audio generation in one call
- [Voice Changer (Quote)](https://docs.venice.ai/api-reference/endpoint/audio/voice-changer/quote): Get a price quote for converting a source recording
- [Voice Changer (Queue)](https://docs.venice.ai/api-reference/endpoint/audio/voice-changer/queue): Queue a speech-to-speech conversion (charged at queue time; do not retry)
- [Voice Changer (Retrieve)](https://docs.venice.ai/api-reference/endpoint/audio/voice-changer/retrieve): Poll conversion status and download converted audio
- [Voice Changer (Complete)](https://docs.venice.ai/api-reference/endpoint/audio/voice-changer/complete): Release stored media after downloading converted audio

### Video
- [Video Generation (Queue)](https://docs.venice.ai/api-reference/endpoint/video/queue): Queue video generation (text-to-video or image-to-video)
- [Video Generation (Retrieve)](https://docs.venice.ai/api-reference/endpoint/video/retrieve): Retrieve generated video by ID
- [Video Generation (Quote)](https://docs.venice.ai/api-reference/endpoint/video/quote): Get price quote for video generation
- [Video Generation (Complete)](https://docs.venice.ai/api-reference/endpoint/video/complete): Queue and wait for video generation in one call

### Tools
- [Text Parser](https://docs.venice.ai/api-reference/endpoint/augment/text-parser): Extract text from PDF, DOCX, XLSX, and plain text files. Runs in-memory on Venice infrastructure with zero data retention ($0.01/request)
- [Web Scrape](https://docs.venice.ai/api-reference/endpoint/augment/scrape): Scrape a web page and return its content as markdown ($0.01/request)
- [Web Search](https://docs.venice.ai/api-reference/endpoint/augment/search): Search the web with privacy-preserving providers — Brave (ZDR, zero data retention) or Google (proxied through Venice so your identity is not associated with the search) ($0.01/request)
- [Crypto Networks](https://docs.venice.ai/api-reference/endpoint/crypto/networks): List all supported blockchain networks (public endpoint, no auth required)
- [Crypto RPC](https://docs.venice.ai/api-reference/endpoint/crypto/rpc): **Venice provides blockchain RPC access** — send JSON-RPC requests to Ethereum, Base, Arbitrum, Optimism, Polygon, Linea, Avalanche, BSC, Blast, zkSync Era, Starknet, Solana, and Robinhood Chain (mainnet + testnets). One API key, unified billing in Venice credits. Supports batch requests (up to 100), idempotent retries, and x402 wallet auth. No separate RPC provider signup needed
- [Crypto RPC Postman Collection](https://www.postman.com/veniceai/workspace/venice-ai-workspace/folder/38652128-2cf5a817-41cd-438b-ad37-5d07c3f13005?action=share&creator=48156591&active-environment=38652128-ef110f4e-d3e1-43b5-8029-4d6877e62041): 27 ready-to-run examples for crypto RPC calls

### Embeddings
- [Embeddings](https://docs.venice.ai/api-reference/endpoint/embeddings/generate): Generate vector embeddings for semantic search

### Decisions (Beta)
- [Create Decisions](https://docs.venice.ai/api-reference/endpoint/decisions/create): `POST /decisions` evaluates a `state` against typed `questions` with the `jev-latest` model and returns structured answers instead of generated text. Question types: `noul` (yes/no probability), `choice` (one option plus a probability distribution), and `score` (probability-weighted position on an ordered rubric). `questions` must contain at least one entry. Beta: schemas and behavior may change
- [System One Compatibility](https://docs.venice.ai/api-reference/endpoint/decisions/systemone): `POST /systemone` is an alias of `POST /decisions` for drop-in compatibility with TypeSafe SDKs (beta)

### Models
- [List Models](https://docs.venice.ai/api-reference/endpoint/models/list): Get available models with capabilities and pricing
- [Model Traits](https://docs.venice.ai/api-reference/endpoint/models/traits): Map trait names to current model IDs per `?type=` (e.g., `default`, `default_code`, `most_uncensored` for text; `fastest`, `highest_quality` for images)
- [Compatibility Mapping](https://docs.venice.ai/api-reference/endpoint/models/compatibility_mapping): Map OpenAI model names to Venice equivalents

### Characters
- [List Characters](https://docs.venice.ai/api-reference/endpoint/characters/list): List available AI character personas
- [Get Character](https://docs.venice.ai/api-reference/endpoint/characters/get): Get details for a specific character
- [Character Reviews](https://docs.venice.ai/api-reference/endpoint/characters/reviews): Get reviews for a character

### API Keys
- [List API Keys](https://docs.venice.ai/api-reference/endpoint/api_keys/list): List your API keys
- [Get API Key](https://docs.venice.ai/api-reference/endpoint/api_keys/get): Get details for a specific API key
- [Create API Key](https://docs.venice.ai/api-reference/endpoint/api_keys/create): Create a new API key
- [Update API Key](https://docs.venice.ai/api-reference/endpoint/api_keys/update): Update an existing API key
- [Delete API Key](https://docs.venice.ai/api-reference/endpoint/api_keys/delete): Delete an API key
- [Generate Web3 Key](https://docs.venice.ai/api-reference/endpoint/api_keys/generate_web3_key/post): Generate API key via Web3 wallet signature
- [Rate Limits](https://docs.venice.ai/api-reference/endpoint/api_keys/rate_limits): Get rate limit configuration for an API key
- [Rate Limit Logs](https://docs.venice.ai/api-reference/endpoint/api_keys/rate_limit_logs): Get rate limit event logs

### Billing
- [Balance](https://docs.venice.ai/api-reference/endpoint/billing/balance): Get current account balance
- [Usage History](https://docs.venice.ai/api-reference/endpoint/billing/usage-history): Get per-request usage history (keyset-paginated)
- [Usage (removed)](https://docs.venice.ai/api-reference/endpoint/billing/usage): `GET /billing/usage` returns `410`; use Usage History instead
- [Usage Analytics](https://docs.venice.ai/api-reference/endpoint/billing/usage-analytics): Get detailed usage analytics

### x402 Wallet Authentication
- [x402 Balance](https://docs.venice.ai/api-reference/endpoint/x402/balance): Get wallet credit balance
- [x402 Top-Up](https://docs.venice.ai/api-reference/endpoint/x402/top-up): Add USDC credits from a Base or Solana wallet
- [x402 Transactions](https://docs.venice.ai/api-reference/endpoint/x402/transactions): Get wallet transaction history

### Reference
- [Rate Limiting](https://docs.venice.ai/api-reference/rate-limiting): Rate limits and best practices
- [Error Codes](https://docs.venice.ai/api-reference/error-codes): API error reference

## Models

- [Model Overview](https://docs.venice.ai/models/overview): Browse all available models
- [Text Models](https://docs.venice.ai/models/text): LLMs including venice-uncensored-1-2, GLM-4.7, Qwen3, Mistral, Llama, Grok, DeepSeek, Kimi
- [Image Models](https://docs.venice.ai/models/image): Diffusion models for image generation (Flux, Stable Diffusion)
- [Text-to-Speech Models](https://docs.venice.ai/models/text-to-speech): TTS models with multilingual voice support (Kokoro, Qwen 3 TTS)
- [Speech-to-Text Models](https://docs.venice.ai/models/speech-to-text): Audio transcription models (Whisper, Parakeet)
- [Music Models](https://docs.venice.ai/models/music): Music and sound effects generation with lyrics support
- [Video Models](https://docs.venice.ai/models/video): Text-to-video and image-to-video models
- [Embedding Models](https://docs.venice.ai/models/embeddings): Vector embedding models

## Guides

### Text & Chat
- [Typed Decisions with Jev (Beta)](https://docs.venice.ai/guides/features/decisions): Classify, route, and score application state with `POST /decisions` and `jev-latest`, and act on the returned probabilities and confidence. Use it for bounded judgments instead of prompting a chat model for JSON
- [Structured Responses](https://docs.venice.ai/guides/features/structured-responses): Get JSON responses with guaranteed schemas using `response_format`
- [Reasoning Models](https://docs.venice.ai/guides/features/reasoning-models): Use models with step-by-step reasoning (supports `reasoning_effort` parameter)
- [Function Calling](https://docs.venice.ai/guides/features/function-calling): Let Venice chat models call your tools with OpenAI-compatible function calling
- [Building a Tool-Using Agent with Function Calling](https://docs.venice.ai/guides/features/tool-using-agent): Run the full tool-calling loop against a database the model has never seen, including parallel calls, errors it recovers from, and guards it cannot talk past
- [Characters](https://docs.venice.ai/guides/features/characters): Discover public Venice characters and chat with their personas via chat completions
- [Vision](https://docs.venice.ai/guides/features/vision): Analyze images with Venice vision models using multimodal content in chat completions
- [File Inputs](https://docs.venice.ai/guides/features/file-inputs): Attach PDF, Office, text, data, and source-code files directly to chat completions — Venice extracts them to text before inference
- [Prompt Caching](https://docs.venice.ai/guides/features/prompt-caching): Reduce latency and costs with prompt caching
- [TEE & E2EE Models](https://docs.venice.ai/guides/features/tee-e2ee-models): Privacy-enhanced AI with Trusted Execution Environments (TEE) and End-to-End Encryption (E2EE); verify with `GET /tee/attestation` and `GET /tee/signature`

### Image
- [Image Generation Guide](https://docs.venice.ai/guides/media/image-generation): Best practices for image generation
- [Prompt Enhancement](https://docs.venice.ai/guides/media/prompt-enhancement): Rewrite image prompts automatically with Venice's enhance_prompt parameter
- [Image Editing Guide](https://docs.venice.ai/guides/media/image-editing): Image editing and inpainting techniques
- [Transparent Background Images](https://docs.venice.ai/guides/media/transparent-backgrounds): Generate and edit PNG/WebP images with a real alpha channel via transparent_background
- [Image Upscaling](https://docs.venice.ai/guides/media/image-upscaling): Enhance and upscale images with Venice's synchronous /image/upscale endpoint

### Video
- [Video Generation Guide](https://docs.venice.ai/guides/media/video-generation): Video generation best practices
- [Seedance 2.0 & 2.5](https://docs.venice.ai/guides/media/seedance-2-0): Generate, edit, extend, and stitch videos with Seedance 2.0 and 2.5 on Venice
- [Seedance Face-Media Consent](https://docs.venice.ai/guides/media/seedance-face-consent): Submit face-bearing media to Seedance 2.0 with the two-call attestation flow
- [Reference to Video](https://docs.venice.ai/guides/media/reference-to-video): Lock in characters, objects, and scenes across AI video generations using reference images on Kling O3 and Grok Imagine R2V
- [Video Upscaling](https://docs.venice.ai/guides/media/video-upscaling): Enhance existing videos to higher resolution (2x/4x) or quality using the Topaz Video Upscale model

### Audio
- [Text-to-Speech](https://docs.venice.ai/guides/media/text-to-speech): Generate spoken audio from text with Venice text-to-speech models and /audio/speech
- [Narrating Articles with Text-to-Speech](https://docs.venice.ai/guides/media/article-narration): Chunk long text past the 4096-character cap, synthesize it, and join the pieces into one narrated audio file
- [Speech-to-Text](https://docs.venice.ai/guides/media/speech-to-text): Transcribe audio to text with Venice speech-to-text models and /audio/transcriptions
- [Meeting Notes with Speech to Text](https://docs.venice.ai/guides/media/meeting-notes): Transcribe a recording with segment timings, extract decisions and owners against a schema, and work around the absence of speaker labels
- [Voice Cloning](https://docs.venice.ai/guides/media/voice-cloning): Clone a voice from a short reference sample with Chatterbox HD via `POST /audio/voices`, then generate speech with Venice text-to-speech
- [Voice Changer](https://docs.venice.ai/guides/media/voice-changer): Convert a source recording into another voice with `/audio/voice-changer/*` — not TTS, not voice cloning, and not `/audio/queue`
- [Music & Sound Effects](https://docs.venice.ai/guides/media/music-and-sound-effects): Generate music and sound effects with Venice's asynchronous audio API

### Search & RAG
- [Embeddings](https://docs.venice.ai/guides/features/embeddings): Generate vector embeddings for semantic search, retrieval-augmented generation, clustering, and recommendations
- [Document Processing](https://docs.venice.ai/guides/tools/document-processing): Extract text and token counts from documents for prompts, embeddings, and retrieval pipelines
- [Web Search and Scraping](https://docs.venice.ai/guides/tools/web-retrieval): Search and scrape the web for fresh context, citations, and retrieval workflows
- [Cited Answers with Web Search](https://docs.venice.ai/guides/tools/cited-web-answers): Chain web search, source selection, scraping, and chat completions into answers where every claim resolves to a sourc

Measured from the saved file

Linked document

UTF-8 text size
26.6 KB27247 bytes
Token count
6696o200k_base encoding
Markdown links
152149 distinct URLs
Fenced code blocks
0243 lines in the file

Size measures the captured UTF-8 text, not compressed transfer size. Tokens are counted with a named tokenizer; other models can count differently. Neither measure establishes context compatibility or answer quality.

Project H1
Venice API

Opening blockquote

Venice is a privacy-first, uncensored AI API platform offering text generation, image generation, audio synthesis, speech-to-speech voice conversion, video generation, music, embeddings, and developer tools (web search, document parsing, blockchain RPC) with zero data retention and OpenAI SDK compatibility.

Context architecture

How the content is divided

7 H2 sections organize 152 parsed Markdown links. 6 links point to .md or .markdown paths; 1 link occurrences use another hostname. Linked destinations have not been recursively validated.

H2 sectionLinksUTF-8 bytesTokens
Getting StartedLine 115614161
API ReferenceLine 195793342257
ModelsLine 1048902247
GuidesLine 11560101912569
LearnLine 19592162486
Key FeaturesLine 20701229301
OptionalLine 232 · Optional101147313

Link counts exclude parsed fenced and inline code and resolve relative destinations against the final endpoint. Section sizes cover the text below each H2 until the next H2.

Embedded material

Code, installation and API examples

Fenced blocks
0 closed
Declared languages
No fenced blocks detected
Package-install commands
0 detected in fenced or inline code
Authentication examples
0 fenced lines match API-key, Authorization, Bearer, OAuth or access-token patterns
HTTP request examples
0 fenced lines match request-method patterns
Authentication links
6 links have authentication-related labels

Embedded example text and links to further documentation are counted separately. A code block does not establish that a sample runs or that an authentication flow succeeds.

Publisher prioritization

The Optional section

1 H2 section named “Optional” contain 10 parsed links. 142 links sit outside those sections. That puts 6.6% of the file’s parsed links under Optional.

The community proposal uses Optional by convention for secondary links that can be skipped for a shorter context. This label does not establish that everything outside it is critical, or that the material inside it is deprecated.

Machine-readable routes

MCP references and configuration text

MCP-related text was detected in the captured file. 1 links have MCP-related labels or paths. 0 fenced lines match mcpServers, mcp_servers or an mcp add command.

These are text and link observations. No MCP server connection, authentication or tool call was tested.

Index and bundle

llms-full.txt companion

Verified text companion

Text companion verified

https://docs.venice.ai/llms-full.txt ↗
Discovery
Linked from this llms.txt
HTTP response
200 · text/plain; charset=utf-8
Text size
1.1 MB · 1204748 bytes (measured)
Token count
311434 o200k_base tokens
Retrieval scope
Complete body retrieved

A verified text response establishes a companion at the checked URL. It does not prove the bundle contains all product documentation, that generation is automated, or that it fits an assistant’s context window.

Change evidence

What can be said about freshness

This capture establishes a baseline. Update frequency cannot be determined from one observation.

Linked release and migration material

2 links have release, changelog, deprecation or migration terms in their labels or paths. These provide places to investigate product changes separately.

File changes can reveal documentation edits or navigation changes. They do not by themselves measure product release velocity, accuracy, agent adoption or AI visibility.

Capture identity and measurement method
Requested endpoint
https://docs.venice.ai/llms.txt
Final endpoint
https://docs.venice.ai/llms.txt
HTTP status
200
Content type
text/plain; charset=utf-8
Text retrieval
Firecrawl literal text capture
Tokenizer
tiktoken 0.14.0 · o200k_base
SHA-256 of saved UTF-8 text
72e00438b2e41c692673f630916dd7caaa8b1eae1bbf15244a5763e5bdc27c80

Counts use a deterministic parser, not a spec-compliance grade. Unusual Markdown and indented code can fall outside its recognized patterns. See the llms.txt proposal and tokenizer implementation.

Related tools and resources

Contact Jon