AI Pricing Hub

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

Source-backed AI model pricing, rankings, history, and benchmark data.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "AI Pricing Hub": {
      "url": "https://aipricinghub.com/mcp/rpc",
      "transport": "streamable-http"
    }
  }
}

Measured tools (11)

• search_models read-only

  Search the pricing catalog with provider, workload, context, and price filters.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • query — ['string', 'number', 'array'] — query
  • cursor — ['string', 'number', 'array'] — cursor
  • provider — ['string', 'number', 'array'] — provider
  • workload — ['string', 'number', 'array'] — workload
  • max_combined_price — ['string', 'number', 'array'] — max combined price
  • min_context_tokens — ['string', 'number', 'array'] — min context tokens

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• compare_models read-only

  Compare pricing, context, and sourced benchmarks for model IDs.

  Parameters:

  • model_ids — ['string', 'number', 'array'] — model ids

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• calculate_cost read-only

  Estimate cost from total input tokens, their cached subset, output tokens and request count. Optional batch_mode uses available batch rates with standard-rate fallback.

  Parameters:

  • model_id — ['string', 'number', 'array'] — model id
  • requests — ['string', 'number', 'array'] — requests
  • batch_mode — boolean — batch mode
  • input_tokens — ['string', 'number', 'array'] — input tokens
  • output_tokens — ['string', 'number', 'array'] — output tokens
  • cached_input_tokens — ['string', 'number', 'array'] — cached input tokens

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• find_cheapest read-only

  Find cheapest models by combined listed input plus output token price.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • cursor — ['string', 'number', 'array'] — cursor
  • provider — ['string', 'number', 'array'] — provider
  • workload — ['string', 'number', 'array'] — workload

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• find_best_value read-only

  Rank models by available benchmark signal per listed token price.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • cursor — ['string', 'number', 'array'] — cursor
  • provider — ['string', 'number', 'array'] — provider
  • benchmark_id — ['string', 'number', 'array'] — benchmark id

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• provider_summary read-only

  Summarize model count, prices, benchmarks, and changes for a provider.

  Parameters:

  • provider — ['string', 'number', 'array'] — provider

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• pricing_history read-only

  Return historical pricing snapshots for a model ID.

  Parameters:

  • model_id — ['string', 'number', 'array'] — model id

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• benchmark_lookup read-only

  Find sourced benchmark rows by model, provider, or benchmark ID.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • cursor — ['string', 'number', 'array'] — cursor
  • model_id — ['string', 'number', 'array'] — model id
  • provider — ['string', 'number', 'array'] — provider
  • benchmark_id — ['string', 'number', 'array'] — benchmark id

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• latest_changes read-only

  Return recent model launches, removals, and pricing changes.

  Parameters:

  • type — ['string', 'number', 'array'] — type
  • limit — ['string', 'number', 'array'] — limit
  • cursor — ['string', 'number', 'array'] — cursor
  • provider — ['string', 'number', 'array'] — provider

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• search_news read-only

  Search AI Pricing Hub news items.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • query — ['string', 'number', 'array'] — query
  • cursor — ['string', 'number', 'array'] — cursor
  • source — ['string', 'number', 'array'] — source
  • language — ['string', 'number', 'array'] — language

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

• recommend_model read-only

  Recommend models for a workload using price, context, and benchmark evidence.

  Parameters:

  • limit — ['string', 'number', 'array'] — limit
  • budget — ['string', 'number', 'array'] — budget
  • provider — ['string', 'number', 'array'] — provider
  • workload — ['string', 'number', 'array'] — workload
  • min_context_tokens — ['string', 'number', 'array'] — min context tokens

  Hints: read-only yes · destructive not declared · idempotent not declared · open-world not declared

Publisher website Repository

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.