# Lovely Router docs

> One prepaid API key for Claude, GPT and Gemini (native Anthropic, OpenAI and Gemini APIs at https://lovelyrouter.com), plus pay-per-result data tools and a browser agent API. Every docs page below is plain Markdown.

Every page of the Lovely Router docs as plain Markdown. All of them in one file: https://lovelyrouter.com/docs/llms-full.txt. Product pages and guides: https://lovelyrouter.com/llms.txt.

## Get started

- [Overview](https://lovelyrouter.com/docs/index.md): One prepaid API key for Claude, GPT and Gemini through their native APIs, plus pay-per-result data tools and a browser agent your code can call.
- [Quickstart](https://lovelyrouter.com/docs/quickstart.md): Create an API key, make your first request with the SDK you already use, and check what it cost.
- [Authentication](https://lovelyrouter.com/docs/authentication.md): Authenticate every request with a Lovely Router API key, sent in whichever header your SDK already uses.
- [Use with your tools](https://lovelyrouter.com/docs/integrations.md): Point Claude Code, Codex CLI, Gemini CLI, desktop chat apps and the official SDKs at Lovely Router with one API key.

## Playground

- [Overview](https://lovelyrouter.com/docs/playground.md): Try chat, image, video, music and speech models from the dashboard without writing code, billed and logged like API calls.
- [Chat](https://lovelyrouter.com/docs/playground/chat.md): Talk to any chat model from the dashboard, tune the system prompt, temperature and output limit, and see the speed and cost of every reply.
- [Images](https://lovelyrouter.com/docs/playground/images.md): Generate and edit pictures from the dashboard, combine up to 16 reference images, keep a saved character's face across images, and download or re-edit the results.
- [Create video](https://lovelyrouter.com/docs/playground/create-video.md): Cast saved characters and places, write each shot with @tags and dialogue, render the shots and join them into one video with sound.
- [Character swap](https://lovelyrouter.com/docs/playground/character-swap.md): Swap people in a video for your characters, shot by shot, keeping the motion, the camera and the voices.
- [Live video](https://lovelyrouter.com/docs/playground/live-video.md): Start an endless video in the playground, give it a first frame, a director and a voice, and change what happens next while it plays.
- [Music video](https://lovelyrouter.com/docs/playground/music-video.md): Upload a song, pick the part to film, write a short brief, and get a music video cut on the beat with the singer lip-syncing every line.
- [Speech](https://lovelyrouter.com/docs/playground/speech.md): Turn text into speech in the playground with preset voices, voices you describe, or your own cloned voice, and download the clips.
- [Talking character](https://lovelyrouter.com/docs/playground/talking-character.md): Bring a saved character to life in the playground — rig it once from its picture, then make it talk in any voice, change expression and move to a song.
- [Characters & scenes](https://lovelyrouter.com/docs/playground/library.md): Save the people and places you use again, from a photo, a picture or a description, and use them in the image and video playgrounds.

## Integrations

- [Claude Code](https://lovelyrouter.com/docs/integrations/claude-code.md): Use Claude Code with a Lovely Router API key: set the base URL and key, pick Claude models, and keep prompt caching working.
- [Codex CLI](https://lovelyrouter.com/docs/integrations/codex-cli.md): Use Codex CLI with a Lovely Router API key: add it as a model provider in config.toml, pick a GPT model, and keep prompt caching working.
- [Gemini CLI](https://lovelyrouter.com/docs/integrations/gemini-cli.md): Use Gemini CLI with a Lovely Router API key: set the base URL and key in two environment variables and pick a Gemini model.
- [Cherry Studio](https://lovelyrouter.com/docs/integrations/cherry-studio.md): Use Cherry Studio with a Lovely Router API key: add it as an OpenAI-compatible provider in one click or by hand, then add Claude, GPT and Gemini models.
- [DeepChat](https://lovelyrouter.com/docs/integrations/deepchat.md): Use DeepChat with a Lovely Router API key: add it as an OpenAI-compatible provider in one click or by hand, then add Claude, GPT and Gemini models.
- [AionUI](https://lovelyrouter.com/docs/integrations/aionui.md): Use AionUI with a Lovely Router API key: add it as an OpenAI-compatible provider in one click or by hand, then add Claude, GPT and Gemini models.
- [OpenAI-compatible apps](https://lovelyrouter.com/docs/integrations/openai-compatible.md): Use any OpenAI-compatible app, agent framework or library with a Lovely Router API key: set the base URL, the key and a model id.
- [SDKs](https://lovelyrouter.com/docs/integrations/sdks.md): Use the official OpenAI, Anthropic and Google Gen AI SDKs with a Lovely Router API key: pass the base URL and key when you create the client.

## Models API

- [How the gateway works](https://lovelyrouter.com/docs/models.md): Lovely Router passes each request to a model in the API format you sent it. This page covers the endpoints, model ids, auth headers, and the few things the gateway changes.
- [Anthropic Messages](https://lovelyrouter.com/docs/models/anthropic.md): Call Claude through the Anthropic Messages API with the official SDKs. Streaming, extended thinking, tool use, prompt caching and token counting work as documented by Anthropic.
- [OpenAI Chat Completions](https://lovelyrouter.com/docs/models/openai.md): Call GPT, Claude and Gemini models through the OpenAI Chat Completions API with the official OpenAI SDKs. Covers streaming with usage, tools, reasoning models and what is not supported.
- [Gemini](https://lovelyrouter.com/docs/models/gemini.md): Call Gemini models through the native Gemini API with the Google Gen AI SDKs or the Gemini CLI. Covers generateContent, streaming with alt=sse, countTokens and thinking.
- [Images](https://lovelyrouter.com/docs/models/images.md): Generate and edit images with GPT Image 2 and GPT Image 2.5 through the OpenAI Images API and the official OpenAI SDKs, at a flat price per image. Covers sizes, edits, reference images, pricing and limits.
- [Video](https://lovelyrouter.com/docs/models/videos.md): Character swap through the Videos API — send a clip and a picture of a new character, get the same shot back with that character, its motion, framing and voice kept — and video shots made from a prompt with tagged characters and scenes. Requests, polling, pricing, regions and limits.
- [MiniMax H3 terms](https://lovelyrouter.com/docs/models/minimax-h3-license.md): The terms you accept to use MiniMax H3 models on Lovely Router — where they may be used, what they may not be used for, the consent you confirm on every request, and the full MiniMax H3 Community License Agreement.
- [Live video](https://lovelyrouter.com/docs/models/live-video.md): An endless video that keeps going while you watch — POST /v1/videos/live streams ~5 s chunks, each carrying on from the one before, and you change what happens next at any moment. Events, steering, pricing and limits.
- [Music videos](https://lovelyrouter.com/docs/models/music-videos.md): Send a song, get a music video back — POST /v1/videos/music cuts it into scenes on the beat, writes a shot for each from your brief, renders every scene to its part of the song and joins them over the original track. Brief, character profile, review, pricing and timing.
- [Text to speech](https://lovelyrouter.com/docs/models/speech.md): Turn text into speech with Qwen3-TTS and ZONOS2 through OpenAI's speech API — preset voices, voices designed from a description, voice cloning from a short clip, emotions, 10 to 40+ languages — priced per minute of audio.
- [LTX terms](https://lovelyrouter.com/docs/models/ltx-license.md): The terms you accept to use LTX models on Lovely Router — what they may not be used for, the consent you confirm on every request, and the full LTX-2.x Community License Agreement.
- [Character rigs](https://lovelyrouter.com/docs/models/character-rig.md): Turn one picture of a character into a layered, rigged 2.5D character that turns its head, blinks, talks with lip sync, changes expression and moves to music — one flat price per character, downloaded as a .livliv package.
- [Character rig terms](https://lovelyrouter.com/docs/models/character-rig-license.md): The terms you accept to make character rigs on Lovely Router — the use restrictions of the CreativeML Open RAIL licences of the model, the consent you confirm on every request, and the credits.
- [Hosted models](https://lovelyrouter.com/docs/models/hosted.md): Open-weight models Lovely Router runs on its own dedicated GPUs, in the Creative · Hosted group. They sleep when idle, so the first request after a quiet spell takes a little longer.
- [Models & pricing](https://lovelyrouter.com/docs/models/pricing.md): List the available models and their per-token prices with GET /v1/models, and see exactly how the charge for each call is computed.
- [Channel groups](https://lovelyrouter.com/docs/models/channel-groups.md): A channel group is the pool of capacity that serves a model, with its own price multiplier. Bind a key to a group or choose one per request, and see which group served each call.
- [Prompt caching](https://lovelyrouter.com/docs/models/caching.md): Pay a fraction of the input price for repeated prompt prefixes. Use each vendor's caching, and let the gateway keep every conversation on the channel that holds its cache.
- [Streaming](https://lovelyrouter.com/docs/models/streaming.md): Stream tokens as they are generated in all three API formats. Learn how the gateway hides early failures, how stalls and mid-stream errors appear, and what a cancelled stream costs.
- [Rate limits & failover](https://lovelyrouter.com/docs/models/rate-limits.md): How per-key limits, capacity queueing and automatic failover work, which status codes they return, and how to retry safely.

## Data tools

- [Overview](https://lovelyrouter.com/docs/tools.md): Pay-per-result scrapers, search and enrichment tools your code or agent can call with the same key and balance as the Models API.
- [Find the right tool](https://lovelyrouter.com/docs/tools/find.md): Browse the catalog, search it by keyword, describe a task to get an ordered plan with costs, and read a tool's input schema before you run it.
- [Run a tool](https://lovelyrouter.com/docs/tools/run.md): Start a run, get results in one call or poll for them, page through large result sets, stop a run early, and see exactly what it cost.
- [MCP server](https://lovelyrouter.com/docs/tools/mcp.md): Connect Claude Code, Cursor, Claude Desktop or any MCP client to the data tools and the browser agent with one URL and your API key.

## Browser agent

- [Overview](https://lovelyrouter.com/docs/browser-agent.md): Send a goal and a start URL; an AI agent does the task in a real Chromium browser and returns an answer and structured JSON.
- [Live events & screen](https://lovelyrouter.com/docs/browser-agent/live.md): Stream what the agent is doing, each step as it finishes and the browser's screen while a run works, instead of waiting for the result.
- [Modes, models & pricing](https://lovelyrouter.com/docs/browser-agent/modes.md): Choose fast or accurate mode, optionally run the agent on a model you pick, and see exactly how a run is reserved and charged.
- [Structured output](https://lovelyrouter.com/docs/browser-agent/structured-output.md): Pass an output schema and get validated JSON in data, built only from the pages the agent visited.
- [Human in the loop](https://lovelyrouter.com/docs/browser-agent/human-in-the-loop.md): Let the agent ask you questions, wait for your approval before it submits, hand you the browser for a CAPTCHA, and let you take over at any time.
- [Logins, secrets & profiles](https://lovelyrouter.com/docs/browser-agent/logins.md): Store write-only secrets the agent can type but never see, generate two-factor codes, and save sessions so later runs start signed in.
- [File uploads](https://lovelyrouter.com/docs/browser-agent/files.md): Upload files once, let a run use them, and the agent attaches the right one to each upload field.
- [Webhooks](https://lovelyrouter.com/docs/browser-agent/webhooks.md): Get an HTTPS POST when a browser agent run starts waiting for a person or finishes, signed so you can verify it came from Lovely Router.

## Cloud phone

- [Cloud phones](https://lovelyrouter.com/docs/phone.md): Real Android 14 phones in the cloud for your agents. Install any APK, let our phone agent work a goal in plain words, or drive the phone yourself with taps, typing and screenshots.

## Cloud agent

- [Cloud agents](https://lovelyrouter.com/docs/agents.md): Lovely Agent in private cloud sandboxes. Create as many agents as you need, for yourself or for your own users, use them in the browser or over the API, share the pages and files they make, and pay per second while they run.
- [White label](https://lovelyrouter.com/docs/agents/white-label.md): Put your own name, logo, colors and fonts on your agents' pages, so your users see your product and nothing that names Lovely Router. Save several brandings and pick one for each agent.

## Account

- [API keys](https://lovelyrouter.com/docs/account/api-keys.md): Create keys, limit what each one can spend and call, and rotate or revoke them without downtime.
- [Credits & billing](https://lovelyrouter.com/docs/account/billing.md): How your prepaid balance works, how to add credit, and how model requests, tool runs and browser runs are charged.
- [Usage & logs](https://lovelyrouter.com/docs/account/usage.md): Track spend, tokens, latency and errors across your keys, inspect any request, and reconcile what you were charged.
- [Account security](https://lovelyrouter.com/docs/account/security.md): Protect your dashboard with two-factor authentication, manage your password and sessions, and know what Lovely Router stores.

## API reference

- [Introduction](https://lovelyrouter.com/docs/api.md): Base URLs, authentication, request conventions, limits and retry guidance that apply to every Lovely Router endpoint.
- [Models endpoints](https://lovelyrouter.com/docs/api/models.md): Reference for the LLM endpoints in Anthropic, OpenAI and Gemini format, with the Lovely Router headers each one accepts and returns.
- [Tools endpoints](https://lovelyrouter.com/docs/api/tools.md): Reference for every /v1/tools endpoint, the tool object and the run object.
- [Browser endpoints](https://lovelyrouter.com/docs/api/browser.md): Every /v1/browser endpoint for runs, vault secrets, files and profiles, with parameters, limits, responses and errors.
- [Phone endpoints](https://lovelyrouter.com/docs/api/phones.md): Every /v1/phones endpoint for cloud phones, apps, the live screen, direct control and agent runs, with parameters, limits, responses and errors.
- [Music video endpoints](https://lovelyrouter.com/docs/api/music-videos.md): Every /v1/videos/music endpoint — create a music video from a song, poll it, review and edit its scenes, render, download the video, one scene or a look's reference sheet, delete — with parameters, responses and errors.
- [Character endpoints](https://lovelyrouter.com/docs/api/characters.md): Every /v1/characters and /v1/scenes endpoint — save a character (a name, a description and a turnaround sheet) or a scene (a place and its picture), list, retrieve, update and delete them, download their pictures, and use them in music videos, character swaps and video shots.
- [Video shot endpoints](https://lovelyrouter.com/docs/api/video-shots.md): Every /v1/videos/shots, /v1/videos/joins and /v1/videos/projects endpoint — render shots from a prompt with tagged characters and scenes, join finished shots into one video for free, and keep storyboards as projects.
- [Video replacement endpoints](https://lovelyrouter.com/docs/api/video-replacements.md): Every /v1/videos/replacements endpoint — upload a whole video, get its shots and the people in it, map each person to a character, and get the video back with them swapped.
- [Audio endpoints](https://lovelyrouter.com/docs/api/audio.md): POST /v1/audio/speech turns text into speech (OpenAI's request shape plus voice design, cloning, language and emotion), and /v1/audio/voices saves voice samples to clone — create, list, retrieve, rename, download and delete them.
- [Agent endpoints](https://lovelyrouter.com/docs/api/agents.md): Every /v1/agents endpoint for cloud agents: create agents for your users, settings, start and stop, run tasks over the API, files and share links, open the web page, agent tokens, and brandings for white-labeled agents.
- [Errors](https://lovelyrouter.com/docs/api/errors.md): Every HTTP status Lovely Router returns, the error body for each API format, and when and how to retry.
