# UberLLM

> UberLLM is a paid MCP server for ai models and llm tooling. Price, generate and edit images and videos across AI models with one prepaid API key. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Paid
- **Platforms:** API, CLI
- **Website:** https://uberllm.dev/docs/mcp
- **Repository:** https://github.com/psa-team-source/uberllm
- **MCP Registry name:** dev.uberllm/mcp
- **Categories:** AI models and LLM tooling, design and media
- **Listing updated:** 2026-11-14

## About

UberLLM is a marketplace that sends each request to a third-party inference provider and bills what that provider charged, with no UberLLM fee on requests. The only fee is 5.5% plus $0.50 when you add credit by card, and accounts start at $0. The MCP server lets an agent work with the media side of that API without writing HTTP calls.

The usual video flow is get_model, quote, start_video, then wait_job until the job succeeds. quote prices a request without making anything, and it returns the listed cost plus the maximum that will be held and billed. A max_cost value on any spending tool makes it refuse a request that could cost more, and start_video takes an idempotency key so a retry never pays for a second video. Result links need the same API key to download and files are kept for about 3 hours (image links about 15 minutes).

You can connect and list tools without a key, but every call needs your key as a Bearer token. Calls are subject to the same per-key spend limits as the API, and failed requests cost nothing.

## Features

- **Price before spending:** quote prices an image, video, music or speech request without making it.
- **Spend guard:** max_cost refuses anything that could cost more.
- **Video jobs:** Start a video, then poll or wait for it to finish.
- **Image edit:** Change pictures from a prompt using an editing model.
- **Balance check:** See what you can spend and what is held for running jobs.

## Good for

- Generate an image from a prompt inside a coding agent
- Check the cost of a video before starting it
- Compare models and their live prices
- Check remaining balance before a batch of jobs

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http mcp https://api.uberllm.dev/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://api.uberllm.dev/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "mcp": {
      "url": "https://api.uberllm.dev/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "mcp": {
      "type": "http",
      "url": "https://api.uberllm.dev/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "mcp": {
      "serverUrl": "https://api.uberllm.dev/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "mcp": {
      "type": "http",
      "url": "https://api.uberllm.dev/mcp"
    }
  }
}
```

## Tools

- `list_models`: Models you can call now, with price; filter by chat, image, video or music.
- `get_model`: One model with its price and parameters schema.
- `quote`: Prices an image, video, music or speech request without making anything.
- `generate_image`: Makes pictures from a prompt.
- `edit_image`: Changes pictures following a prompt.
- `start_video`: Starts a video and returns a job.
- `get_job`: Status, progress and file links of a video or music job.
- `wait_job`: Waits until a job ends, up to 50 seconds.
- `balance`: What you can spend now and what is held, in US dollars.

## FAQ

### Do I need an API key?

Yes for tool calls; listing the tools works without one.

### What does UberLLM charge?

Requests cost what the provider charged; adding credit by card adds 5.5% plus $0.50.

### Are failed requests billed?

No.

### How long are results kept?

About 3 hours for job files and about 15 minutes for image links.

### Is UberLLM free?

No. UberLLM is a paid product.

### How do I add UberLLM to Claude Code?

Run this command in your terminal: claude mcp add --transport http mcp https://api.uberllm.dev/mcp

### What are the best alternatives to UberLLM?

The closest alternatives to UberLLM are HyperFrames by HeyGen, ffmpeg-skill, Claude Code Ultimate Guide, AvatarSay — Fictional Character & Talking Video MCP and ChartAnt.

## Alternatives

- [HyperFrames by HeyGen](https://ai-powered.com/mcp/hyperframes-by-heygen): HyperFrames by HeyGen is a freemium MCP server for design and media.
- [ffmpeg-skill](https://ai-powered.com/mcp/ffmpeg-skill): ffmpeg-skill is an open-source MCP server for design and media.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [AvatarSay — Fictional Character & Talking Video MCP](https://ai-powered.com/mcp/avatarsay-fictional-character-and-talking-video-mcp): AvatarSay — Fictional Character & Talking Video MCP is a freemium MCP server for design and media.
- [ChartAnt](https://ai-powered.com/mcp/chartant): ChartAnt is a freemium MCP server for design and media.
- [VfxSmith](https://ai-powered.com/mcp/vfxsmith): VfxSmith is a freemium MCP server for design and media.
- [CalderFlow](https://ai-powered.com/mcp/calderflow): CalderFlow is a free MCP server for design and media.
- [Hourlore](https://ai-powered.com/mcp/hourlore): Hourlore is a freemium MCP server for design and media.

---
Source: https://ai-powered.com/mcp/uberllm
