aiph://providers
providers data from generated AI Pricing Hub datasets.
Model Context Protocol
AI Pricing Hub MCP Server
Connect AI assistants to source-backed AI Pricing Hub providers, models, pricing, benchmarks, history, workloads, news, and changelog data through a read-only MCP endpoint.
Installation
Remote MCP endpoint
Use https://aipricinghub.com/mcp/rpc as the Streamable HTTP JSON-RPC endpoint. The server is read-only, Cloudflare compatible, and backed by generated static datasets.
{
"mcpServers": {
"ai-pricing-hub": {
"url": "https://aipricinghub.com/mcp/rpc"
}
}
}Resources
Context exposed to MCP clients
aiph://models
models data from generated AI Pricing Hub datasets.
aiph://rankings
rankings data from generated AI Pricing Hub datasets.
aiph://benchmarks
benchmarks data from generated AI Pricing Hub datasets.
aiph://history
history data from generated AI Pricing Hub datasets.
aiph://pricing
pricing data from generated AI Pricing Hub datasets.
aiph://calculator
calculator data from generated AI Pricing Hub datasets.
aiph://workloads
workloads data from generated AI Pricing Hub datasets.
aiph://news
news data from generated AI Pricing Hub datasets.
aiph://changelog
changelog data from generated AI Pricing Hub datasets.
aiph://provider-trends
provider trends data from generated AI Pricing Hub datasets.
Tools
Read-only tools
| Name | Purpose |
|---|---|
search_models | Search and filter the model catalog. |
compare_models | Compare price, context, and benchmark evidence for model IDs. |
calculate_cost | Estimate workload cost from input, cached input, output, and request count. |
find_cheapest | Return cheapest models by listed token price. |
find_best_value | Rank models by available benchmark signal per listed token price. |
provider_summary | Summarize provider model counts, prices, and benchmark coverage. |
pricing_history | Return historical pricing snapshots for a model. |
benchmark_lookup | Search sourced benchmark rows. |
latest_changes | List model launches, removals, and pricing changes. |
search_news | Search AI news items. |
recommend_model | Recommend models by workload using source-backed signals. |
Prompts
Reusable assistant workflows
- Best model for codingUses MCP tools and resources without fabricating missing values.
- Best model for RAGUses MCP tools and resources without fabricating missing values.
- Cheapest modelUses MCP tools and resources without fabricating missing values.
- Compare modelsUses MCP tools and resources without fabricating missing values.
- Enterprise recommendationUses MCP tools and resources without fabricating missing values.
Examples
JSON-RPC requests
{
"jsonrpc": "2.0",
"id": 1,
"method": "initialize",
"params": {
"protocolVersion": "2025-06-18",
"capabilities": {},
"clientInfo": {
"name": "example-client",
"version": "1.0.0"
}
}
}{
"jsonrpc": "2.0",
"id": 2,
"method": "tools/call",
"params": {
"name": "find_cheapest",
"arguments": {
"provider": "openai",
"limit": 5
}
}
}Client setup
Cursor, Claude Desktop, OpenAI, and VS Code
Remote MCP server
Add an MCP server named ai-pricing-hub with URL https://aipricinghub.com/mcp/rpc in Cursor MCP settings.
Connector config
Use a remote MCP connector if available, or a local proxy that forwards JSON-RPC requests to https://aipricinghub.com/mcp/rpc.
Agent tools
Register https://aipricinghub.com/mcp/rpc as the MCP server URL in MCP-capable OpenAI agent tooling.
MCP extension
Add the endpoint as a remote MCP server in an MCP-compatible VS Code extension or agent configuration.
Analytics
Request tracking
The MCP function emits response headers for static-compatible analytics: x-aiph-analytics-event: mcp_request, x-aiph-mcp-tool, and x-aiph-mcp-provider. Cloudflare logs or analytics exports can aggregate MCP requests, most used tools, and most requested providers without adding a database.
Continue
Pick up where you left off
Public API
Build with AI Pricing Hub data
Use static JSON endpoints for providers, models, rankings, history, market metrics, and changelog events.
Editorial information
Reviewed by AI Pricing Hub Editorial
2026-08-04
Methodology explains collection, validation, limitations, and update cadence.