Language models
Claude, GPT, and Gemini families for chat, coding, reasoning, and tool use — the same LLM coverage a Requesty user expects, with all the latest model versions.
Requesty is an LLM-only proxy built for Claude Code and Cursor. RunAPI covers the same text routing — 220+ models including Claude, GPT, and Gemini — and adds image, video, music, and audio generation under one key, with an MCP server, a CLI, and SDKs for Python, Node.js, PHP, Java, Ruby, and Go. All at 50% off official rates.
Both are OpenAI-compatible gateways that route requests to multiple model providers using one key. Requesty focuses on LLM text routing for coding tools like Claude Code and Cursor. RunAPI covers the same LLM use cases — and passes through the same model costs at 50% off official rates — while adding image, video, music, and audio generation plus a developer toolkit.
Both expose an OpenAI-compatible API and route Claude, GPT, and Gemini for tools like Claude Code and Cursor. Switching is a base URL and key change.
Requesty is LLM-only. RunAPI adds image, video, music, and audio generation under one key — no separate accounts.
RunAPI ships an MCP server, a CLI, and SDKs for Python, Node.js, PHP, Java, Ruby, and Go, plus a prompt library. These are not included in an LLM-only proxy.
RunAPI bills pay-as-you-go at 50% of official model rates. No subscription. Calculate your exact cost before you send a request.
The table compares the two gateways across what developers actually evaluate when choosing a proxy: modality coverage, model count, developer tooling, and whether you can swap in a new base URL without touching your code.
| Feature | Requesty | RunAPI |
|---|---|---|
| LLM routing | Yes | Yes |
| Image generation | No | Yes |
| Video generation | No | Yes |
| Music and audio | No | Yes |
| Total models | LLM models | 220+ across all modalities |
| MCP server | No | Yes |
| Official SDKs | OpenAI-compatible only | Python, Node.js, PHP, Java, Ruby, Go + OpenAI-compatible |
| OpenAI-compatible API | Yes | Yes |
Requesty routes language models. RunAPI routes language models and generative media models under one key, so one connection covers text, images, video, music, and audio. If your team uses Claude Code or Cursor today, you swap out your normal connection for RunAPI's — and the same key then unlocks every other modality without a separate account.
Claude, GPT, and Gemini families for chat, coding, reasoning, and tool use — the same LLM coverage a Requesty user expects, with all the latest model versions.
Text-to-image and image-edit models for generation, inpainting, and style transfer, callable from the same API and key.
Text-to-video and image-to-video models for short clips and animation, billed pay-as-you-go per task.
Music generation and audio models for soundtracks, voiceovers, and effects — modalities Requesty does not cover.
Beyond the raw API, RunAPI ships tooling that an LLM-only proxy typically skips: a Model Context Protocol server, a command-line tool, and typed language SDKs. Whether you are routing from Claude Code, Cursor, or your own app, the same key covers everything.
Discover models, check pricing, create tasks, and poll results directly from MCP-aware hosts like Claude Code and Cursor — no code required.
Run generation tasks and manage keys from the terminal. Useful for one-off jobs and scripts outside a code editor.
Typed clients for Python, Node.js, PHP, Java, Ruby, and Go. If you use the OpenAI SDK in any other language, it works without changes — just swap the base URL and key.
A searchable catalog of reusable prompt examples across image, video, and audio modalities, so you can build on proven patterns instead of starting from scratch.
RunAPI bills pay-as-you-go at 50% of each provider's official published rate — no subscription tier required. The table shows representative flagship LLM rates so you can calculate your exact bill before you send a request. Image, video, and audio are priced per task.
| Model | Official input /M | Official output /M | RunAPI input /M | RunAPI output /M |
|---|---|---|---|---|
| Claude Sonnet 4.6 | $6.00 | $30.00 | $3.00 | $15.00 |
| Claude Opus 4.7 | $10.00 | $50.00 | $5.00 | $25.00 |
| GPT-5.4 | $2.50 | $15.00 | $1.25 | $7.50 |
| Gemini 2.5 Pro | $1.25 | $10.00 | $0.63 | $5.00 |
RunAPI applies a 50% discount across all providers. Media models are billed per task. Prices verified June 2026.
Sign up at runapi.ai. The free tier requires no credit card.
Go to Dashboard → API Keys, create a key, and save it.
Change your client's base URL to https://runapi.ai/v1 and use the RunAPI key. Your existing OpenAI-compatible code stays the same — the request and response shapes do not change.
Call image, video, or audio models through the same key, or use the MCP server and CLI for non-code workflows.
Yes for LLM routing. Both are OpenAI-compatible, so switching is usually a two-line change — point your base URL at runapi.ai/v1 and swap in your RunAPI key. Your existing request and response code stays the same. RunAPI then adds image, video, music, and audio models that Requesty does not offer, all under the same account.
RunAPI bills at 50% of each provider's official published rate — no subscription required. Sonnet 4.6 runs $3 input and $15 output per million tokens; GPT-5.4 runs $1.25 input and $7.50 output. Requesty charges a 5% markup on top of provider costs. RunAPI's flat half-off rate means you can calculate your exact bill before you send a request.
Yes. Set the Anthropic or OpenAI base URL in Claude Code or Cursor's settings to RunAPI and use your RunAPI key. Both tools keep running without further changes. The same key also gives you access to image, video, and audio models — modalities Requesty does not cover — so you are not managing separate accounts for different tasks.
No. Point your OpenAI-compatible client's base URL at https://runapi.ai/v1 and use your RunAPI key. The request and response shapes are identical to the OpenAI API, so your application logic does not change. You can run both gateways side by side during the transition if you want to move gradually.
RunAPI covers 220+ models across every modality: language models from the Claude, GPT, and Gemini families for chat, coding, reasoning, and tool use, plus image, video, music, and audio generation models. All of them are reachable through one key and one OpenAI-compatible API.
RunAPI adds generative media — image, video, music, and audio — alongside the LLM routing both gateways share. It also ships an MCP server, a command-line tool, and official SDKs for Python, Node.js, PHP, Java, Ruby, and Go, plus a searchable prompt library. These are tooling layers that an LLM-only proxy like Requesty typically does not include.
Both. RunAPI publishes typed SDKs for Python, Node.js, PHP, Java, Ruby, and Go. For every other language, the full OpenAI SDK works without modification — just change the base URL and key. It also ships an MCP server for Claude Code and Cursor workflows, and a command-line tool for scripts and terminal use.
How RunAPI compares to OpenRouter on pricing, models, and media support.
Hosted API vs self-hosted proxy — when to use RunAPI over LiteLLM.
Claude Opus, Sonnet, and Haiku pricing at 50% off Anthropic rates.
Browse all 220+ models with pricing, parameters, and code samples.
Create a free RunAPI account for 220+ models across text, image, video, and audio — OpenAI-compatible, at 50% off official rates.