Ask Claude Desktop, Claude Code, Cursor, Cline, or any MCP client about AI model pricing — and get instant answers from 42 models across 10 providers.
Get current input/output costs per 1M tokens for 42 models from OpenAI, Anthropic, Google, DeepSeek, and more.
Side-by-side comparison of any 2-5 models. Instantly see which is cheapest, which has the largest context window.
Estimate monthly costs based on your usage. Input daily requests and token counts to get real dollar amounts.
Sort all models by price. Filter by provider, tier, or minimum context window. Always up to date.
Add this to your Claude Code MCP settings (run claude mcp add or edit ~/.claude.json):
claude mcp add apipulse --transport http https://getapipulse.com/api/mcp
Or manually add to your ~/.claude.json:
{
"mcpServers": {
"apipulse": {
"type": "http",
"url": "https://getapipulse.com/api/mcp"
}
}
}
Use the APIpulse MCP server in the Claude desktop app (macOS/Windows). Edit your Claude Desktop config:
{
"mcpServers": {
"apipulse": {
"type": "http",
"url": "https://getapipulse.com/api/mcp"
}
}
}
Find the config file: Claude Desktop → Settings → Developer → Edit Config. Restart Claude after saving.
Add to your Cursor MCP settings (Settings → MCP → Add Server or edit .cursor/mcp.json):
{
"mcpServers": {
"apipulse": {
"type": "http",
"url": "https://getapipulse.com/api/mcp"
}
}
}
Cline is the most popular MCP client for VS Code. Add the APIpulse server to your Cline MCP settings:
{
"mcpServers": {
"apipulse": {
"type": "streamableHttp",
"url": "https://getapipulse.com/api/mcp"
}
}
}
Or use the Cline UI: click the ⚙️ gear icon → MCP Servers → Add Server → paste the URL.
Continue supports MCP servers in both VS Code and JetBrains IDEs. Add to your Continue config:
mcpServers:
- name: apipulse
transport: http
url: https://getapipulse.com/api/mcp
Any MCP client that supports the Streamable HTTP transport can connect. Works with Windsurf, Zed, Copilot, and more. Use the endpoint URL:
https://getapipulse.com/api/mcp
Protocol: MCP Streamable HTTP (2025-03-26). Supports: initialize, tools/list, tools/call, ping.
Once connected, your AI assistant can call these tools:
| Tool | Description | Parameters |
|---|---|---|
| get_pricing | Get pricing for all or specific models. Filter by provider or tier. | model?, provider?, tier? |
| compare_models | Compare 2-5 models side by side with cost and context window. | models[] (required) |
| calculate_cost | Estimate monthly cost based on usage (requests/day, tokens/request). | model (required), daily_requests?, input_tokens?, output_tokens? |
| find_cheapest | Find cheapest models. Filter by provider, min context, sort order. | provider?, min_context?, sort_by? |
| get_model_details | Detailed info for one model: pricing, context, tier, deprecation status, and replacement recommendation. | model (required) |
Once the MCP server is configured, try asking your AI assistant:
"What's the cheapest AI model with 1M context?" "Compare GPT-5 vs Claude Opus 4.8 pricing" "How much would 10K daily requests cost on Gemini 3 Flash?" "Which provider has the best budget models?" "Calculate my monthly cost: 5K requests/day, 800 input + 300 output tokens on Claude Sonnet 4.6" "Get details on Claude 4 Opus — is it deprecated? What should I use instead?"
No install needed — test the MCP server right here in your browser. Click an example or build your own query:
Is this free?
Yes, completely free. No API key required. Rate limited to 60 requests/minute.
How current is the pricing data?
Updated regularly. Each response includes a timestamp. Check /data/pricing.json for the raw data.
Does it work offline?
No — it's a hosted HTTP endpoint. Your AI tool needs internet access.
What about Pro features?
The MCP server gives you pricing data. APIpulse Pro ($29) adds migration code, cost optimization reports, and provider-specific guides.
Share this with other developers
APIpulse Pro includes migration code templates, cost optimization reports, and provider-specific guides for all 42 models.
Get Pro — $29 Lifetime Try Free Tools