MCP Server

Connect AI coding tools to AICredits with the official MCP server. Live model lists, INR pricing, credit balance, docs search, and test chats — without leaving your editor.

The AICredits MCP (Model Context Protocol) server plugs AICredits into the AI tools you already use. Once connected, your assistant can pull live AICredits data — models, INR pricing, your credit balance, usage, and docs — and send quick test messages, all without leaving your editor.

It's a remote server hosted by AICredits — nothing is installed locally.

Connect

Add the server to your MCP client with your AICredits API key (create one here):

claude mcp add --transport http aicredits https://api.aicredits.in/mcp \
  --header "Authorization: Bearer sk-YOUR-KEY"

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "aicredits": {
      "url": "https://api.aicredits.in/mcp",
      "headers": { "Authorization": "Bearer sk-YOUR-KEY" }
    }
  }
}

Any client supporting streamable HTTP transport:

{
  "url": "https://api.aicredits.in/mcp",
  "transport": "http",
  "headers": { "Authorization": "Bearer sk-YOUR-KEY" }
}

Use a dedicated API key with a budget limit for MCP access — you can set per-key budgets on the Keys page.

Tools

All tools are read-only lookups against live AICredits data, except chat_send, which makes a billable inference call.

Tool	What it does
`models_list`	List available models with live INR pricing per 1M tokens; filter with `search`
`model_get`	Full detail for one model: pricing, context window, modalities, status
`pricing_estimate`	Estimate a request's INR cost from token counts
`credits_balance`	Your current wallet balance in INR
`usage_summary`	Recent request count, INR spend, and top models (up to 90 days)
`docs_search`	Search this documentation and get links to relevant pages
`chat_send`	Send a test message to any model through your account — billable, deducts credits like any API request

Example prompts

Once connected, ask your assistant things like:

"Which is the cheapest model on AICredits that handles 128K context?"
"What's my AICredits balance and what did I spend this week?"
"Estimate the cost of 50K input / 2K output tokens on claude-sonnet-4.5"
"Send 'hello' to gpt-4o-mini through AICredits and show me the usage"

Model recommendations are grounded in live data — the assistant reads current pricing and availability instead of guessing from stale training knowledge.

Notes

chat_send runs through the full AICredits pipeline: authentication, rate limits, guardrails, billing, and request logging — identical to a normal API call.
Responses respect your account's data-retention policy.
The server uses MCP streamable HTTP transport; SSE-only legacy clients are not supported.

Connect

Tools

Example prompts

Notes

On this page