# Liquid Inference

> Liquid Inference is a freemium MCP server for ai models and llm tooling. Remote MCP server for an auction exchange where providers bid to answer LLM prompts. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $5 a month
- **Platforms:** API, Web
- **Website:** https://liquidinference.ai
- **MCP Registry name:** ai.liquidinference/liquid
- **Categories:** AI models and LLM tooling
- **Listing updated:** 2026-11-14

## About

Providers post a per-token price for each model they serve, you set requirements such as models, regions, speed and minimum quality, and the lowest offer that meets them answers the request. Through the MCP server an agent can browse models and live prices, preview an answer and buy one. For regular API use, most OpenAI or Anthropic code needs only a new base URL and API key, with model names taken from the exchange's list.

The exchange fixes a price cap before the first token, so a job never costs more than that cap, and unused cap stays in your balance. Each job gets one receipt showing the winning offer, usage and cost. If a provider fails before any output the exchange can try the next offer, but it never switches mid-answer. Prompts are not stored unless you save one in a workload, and a retention setting of none limits offers to providers that promise not to keep requests.

The exchange adds a fee of the larger of 5% of the provider's charge or 1 cent per million tokens. Extras include monthly token reservations from one provider, an Auto Router called liquid.auto, and bring-your-own-key routing.

## Features

- **Request-level auctions:** The cheapest offer meeting your rules answers each request.
- **Price cap:** A cap is fixed before the first token and a job never exceeds it.
- **Your terms:** Choose models, regions, speed and minimum quality.
- **Itemised receipts:** Each job shows the winning offer, usage and cost.
- **Reservations:** Buy a monthly token pool on one model from one provider.
- **Retention control:** Set retention to none to use only providers that do not keep prompts.

## Good for

- Cut inference spend by letting providers compete
- Compare live model prices from an agent
- Preview an answer before buying it
- Cap the cost of each request
- Reserve a fixed monthly token pool

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http liquid https://router.inference.ai.exchange/mcp
```

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

```sh
https://router.inference.ai.exchange/mcp
```

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

```json
{
  "mcpServers": {
    "liquid": {
      "url": "https://router.inference.ai.exchange/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "liquid": {
      "type": "http",
      "url": "https://router.inference.ai.exchange/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "liquid": {
      "serverUrl": "https://router.inference.ai.exchange/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "liquid": {
      "type": "http",
      "url": "https://router.inference.ai.exchange/mcp"
    }
  }
}
```

## FAQ

### What does the exchange charge?

The larger of 5% of the provider's charge and a floor of 1 cent per million tokens; jobs that fail before any output pay no fee.

### Do I need a card to try it?

No. A new workspace gets a starting credit.

### What are the plans?

Lite $5 a month, Plus $15 and Pro $49, each taking a share off every job; Free allows 50 free-model requests a day and Pay as you go has a 5% fee.

### Does it store my prompts?

No, except a prompt you save in a workload.

### Is Liquid Inference free?

Liquid Inference has a free plan, with paid plans for more usage or features. Paid plans start at $5 a month.

### How do I add Liquid Inference to Claude Code?

Run this command in your terminal: claude mcp add --transport http liquid https://router.inference.ai.exchange/mcp

### What are the best alternatives to Liquid Inference?

The closest alternatives to Liquid Inference are Claude Code Ultimate Guide, signafAIde, Toolkien, Lugh and HypePaper.

## Alternatives

- [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.
- [signafAIde](https://ai-powered.com/mcp/signafaide): signafAIde is a freemium MCP server for security.
- [Toolkien](https://ai-powered.com/mcp/toolkien): Toolkien is an open-source MCP server for ai models and llm tooling.
- [Lugh](https://ai-powered.com/mcp/lugh): Lugh is an open-source MCP server for ai models and llm tooling.
- [HypePaper](https://ai-powered.com/mcp/hypepaper): HypePaper is a freemium MCP server for ai models and llm tooling.
- [Agent Data Tools: Compact (find, describe, run)](https://ai-powered.com/mcp/agent-data-tools-compact-find-describe-run): Agent Data Tools: Compact (find, describe, run) is a freemium MCP server for developer tools.
- [Angel Agent Guard: pre-purchase check + prompt-injection scan](https://ai-powered.com/mcp/angel-agent-guard-pre-purchase-check-prompt-injection-scan): Angel Agent Guard: pre-purchase check + prompt-injection scan is a freemium MCP server for ai models and llm tooling.
- [Mizan Calibration](https://ai-powered.com/mcp/mizan-calibration): Mizan Calibration is a free MCP server for ai models and llm tooling.

---
Source: https://ai-powered.com/mcp/liquid-inference
