# Roteo > Roteo is a pay-per-call AI gateway: text (Claude and GPT), image, and video > generation through one API that is both OpenAI- and Anthropic-compatible. You fund a > wallet with a stablecoin and pay per request, with no subscription. This file indexes > the documentation so an LLM or agent can find and read any guide. Machine onboarding in a single file (base URL, how to mint a key, every endpoint, the live model list with exact sizes and durations, and runnable examples): https://api.roteo.ai/v1/llms.txt Human-readable docs are at https://roteo.ai/docs/. Each page's clean Markdown is at https://roteo.ai/docs/raw/ — the links below point at the Markdown so it reads as plain text. ## Documentation - [Overview](https://roteo.ai/docs/raw/overview): One API for Claude, GPT, open-weight, image and video models, paid per request from your own wallet. - [Quickstart](https://roteo.ai/docs/raw/quickstart): Get a key, fund your wallet and make your first request in a couple of minutes. - [Set up in your tools](https://roteo.ai/docs/raw/set-up): Connect Claude Code, the OpenAI SDK, Cursor and image tools to Roteo with one key. - [Models](https://roteo.ai/docs/raw/models): Every model ID you can call, and how to choose between them. - [How it works](https://roteo.ai/docs/raw/how-it-works): How a request travels through Roteo, where the inference comes from, and what Roteo never does. - [Authentication](https://roteo.ai/docs/raw/authentication): Every request is authenticated with a Roteo API key that starts with sk_. - [Pay-per-call billing](https://roteo.ai/docs/raw/billing): Each request is paid onchain from your own wallet, the moment you make it. - [Streaming](https://roteo.ai/docs/raw/streaming): Receive the answer token by token as it is generated, over server-sent events. - [Billing chains](https://roteo.ai/docs/raw/chains): Choose which chain and stablecoin a request settles on. - [Errors](https://roteo.ai/docs/raw/errors): Error shape and the stable status codes the API returns. - [Rate limits](https://roteo.ai/docs/raw/rate-limits): Per-key request limits and how to handle 429s. - [Tool use](https://roteo.ai/docs/raw/tool-use): Let models call your functions and act on the results. - [Reasoning](https://roteo.ai/docs/raw/thinking): Give a model room to reason before it answers, and set how hard it works. - [Station](https://roteo.ai/docs/raw/station): Make images and video in the app, with references, drafts, upscales and a clip editor, your files kept on your device. - [Seedance video](https://roteo.ai/docs/raw/seedance): Make video from a prompt, a starting frame or up to 30 reference images, in the Station or through the API. - [Seedream images](https://roteo.ai/docs/raw/seedream): Generate images from a prompt, or edit and blend up to 10 reference images, in the Station or through the API. - [FAQ](https://roteo.ai/docs/raw/faq): Common questions about wallets, billing and privacy. ## API Reference - [Chat Completions](https://roteo.ai/docs/raw/api/chat-completions): POST /v1/chat/completions: the OpenAI-compatible endpoint. Serves Claude, GPT and open-weight models. - [Messages](https://roteo.ai/docs/raw/api/messages): POST /v1/messages: the Anthropic Messages API, currently serving Claude Code. - [Responses](https://roteo.ai/docs/raw/api/responses): POST /v1/responses: the OpenAI Responses API, for Codex CLI and Responses-mode clients. - [Count tokens](https://roteo.ai/docs/raw/api/count-tokens): POST /v1/messages/count_tokens: count tokens for a prompt before you send it. - [Images](https://roteo.ai/docs/raw/api/images): OpenAI-compatible endpoints for generating and editing images. - [Videos](https://roteo.ai/docs/raw/api/videos): Async, OpenAI-Sora-compatible text-to-video and image-to-video generation. - [Uploads](https://roteo.ai/docs/raw/api/uploads): Send a local image, video or audio file to a video job, past the request size limit. ## SDKs - [OpenAI SDK](https://roteo.ai/docs/raw/sdks/openai): Use the official OpenAI SDKs with Roteo for text and images. - [Anthropic SDK](https://roteo.ai/docs/raw/sdks/anthropic): The Anthropic SDKs call the Messages API, which currently serves Claude Code. ## Integrations - [Agent onboarding](https://roteo.ai/docs/raw/integrations/llms-txt): Point an agent at /v1/llms.txt to wire up Roteo on its own. - [Claude Code](https://roteo.ai/docs/raw/integrations/claude-code): Install Claude Code, point it at Roteo, verify the connection, and run it on Claude or DeepSeek. - [GPT (OpenAI-compatible)](https://roteo.ai/docs/raw/integrations/gpt): Call GPT models through the Chat Completions API with the OpenAI SDK or curl. - [Codex CLI](https://roteo.ai/docs/raw/integrations/codex): Point OpenAI's Codex CLI at Roteo and run GPT models from your wallet. - [Opencode](https://roteo.ai/docs/raw/integrations/opencode): Run Claude, open-weight and GPT models in opencode through Roteo. - [Hermes](https://roteo.ai/docs/raw/integrations/hermes): Run Claude in Hermes with Roteo's plugin, or open-weight and GPT models through its custom provider. - [omp (oh-my-pi)](https://roteo.ai/docs/raw/integrations/omp): Run open-weight, GPT and Claude models in omp through Roteo, all on one key. - [OpenClaw](https://roteo.ai/docs/raw/integrations/openclaw): Onboard OpenClaw with a custom provider pointed at Roteo. - [Cursor](https://roteo.ai/docs/raw/integrations/cursor): Override Cursor's OpenAI base URL to run Roteo models in the editor. - [Agents & OpenAI-compatible clients](https://roteo.ai/docs/raw/integrations/agents): Point any OpenAI-compatible client or agent framework at Roteo. - [Curl](https://roteo.ai/docs/raw/integrations/curl): Copy-paste curl recipes for the Roteo API. ## Referrals - [Referrals](https://roteo.ai/docs/raw/referrals): Earn 10% of everything the people you refer spend on Roteo, paid to your wallet in stablecoins, automatically. ## Roteo Token - [Introduction](https://roteo.ai/docs/raw/token/overview): ROTEO is a discount and yield token backed by Hard Reserves held onchain. - [Tax](https://roteo.ai/docs/raw/token/tax): Buys and sells in taxed ROTEO pools pay a 2% tax, converted to USDG. The Treasury gets 20% of it at launch, rising daily to 100% on day 30; the team gets the rest. - [Bonds](https://roteo.ai/docs/raw/token/bonds): Deposit USDG to receive newly minted ROTEO over 2 days, priced never below Hard NAV. - [Premium sales](https://roteo.ai/docs/raw/token/premium-sales): When ROTEO trades far above Hard NAV, the protocol sells a little new ROTEO and keeps every USDG. - [Staking and yield](https://roteo.ai/docs/raw/token/staking): Stake ROTEO for sROTEO and receive new ROTEO every 8 hours, paid only from reserves above 1 USDG per token, with the Strategy Sleeve left out. - [Lock and discounts](https://roteo.ai/docs/raw/token/lock): Lock ROTEO or wsROTEO for value-based discounts on Roteo products. - [Treasury and Hard NAV](https://roteo.ai/docs/raw/token/treasury): What counts as Hard Reserves, how Hard NAV is calculated, the genesis sale, and where new supply comes from. - [Floor bid](https://roteo.ai/docs/raw/token/floor-bid): A standing bid buys ROTEO just below floor-bid backing and burns it. - [Strategy Sleeve](https://roteo.ai/docs/raw/token/strategy): How Treasury capital is deployed and marked, and why Roteo operations stay outside Hard NAV. - [Contracts](https://roteo.ai/docs/raw/token/contracts): Every contract and operator address the protocol uses on Robinhood Chain. - [Parameters](https://roteo.ai/docs/raw/token/parameters): Every number the protocol runs on, in one place. ## Optional - [Terms of Service](https://roteo.ai/docs/raw/legal/terms): The terms that govern your use of Roteo. - [Privacy Policy](https://roteo.ai/docs/raw/legal/privacy): What Roteo does and does not collect, and how it is used.