MPPtempoquality 0.92

AI-generated answers with web citations via Exa, pay-per-request at $0.01

Price
$0.01 / call
Protocol
mpp
Verified
no

What it does

The Exa Answer endpoint on StableEnrich returns AI-generated answers to natural language questions, grounded in live web search results and accompanied by source citations. It is a POST endpoint at `https://stableenrich.dev/api/exa/answer` that accepts a JSON body with a required `query` string and optional `stream` and `text` boolean flags. The response includes an `answer` string, an array of `citations` (each with URL, title, author, published date, and optionally extracted text), and a `costDollars` object.

This endpoint is part of the StableEnrich platform, which unifies multiple premium data APIs (Exa, Apollo, Firecrawl, Google Maps, Serper, Whitepages, Clado, Minerva, Reddit, Hunter, Influencer, Cloudflare) behind a single micropayment interface. No API keys or subscriptions are required — authentication is handled via HTTP 402 micropayment using x402 or MPP (Tempo method with pathUSD on Tempo L2). The fixed price is $0.01 per request. Payments can be made on Base mainnet (eip155:8453), Solana, or Tempo.

The endpoint is designed for agent workflows using the agentcash MCP server or CLI, which automates payment processing. Agents discover endpoints via `discover_api_endpoints`, check schemas with `check_endpoint_schema`, and execute calls with `fetch`. The Exa Answer endpoint is useful when an agent needs a concise, cited answer to a factual question without manually parsing search results.

Capabilities

ai-answer-generationweb-search-groundedcitation-extractionpay-per-requestmicropayment-authx402-protocolmpp-tempo-settlementstreaming-responseno-api-key-required

Use cases

  • Agent asks a factual question and needs a concise answer with source URLs
  • Automated research pipelines that need cited summaries of web knowledge
  • Building RAG-style workflows where web search provides grounding for LLM answers
  • Quick fact-checking with traceable citations
  • Replacing manual search-then-summarize steps in agent toolchains

Fit

Best for

  • AI agents needing cited answers to factual questions without parsing raw search results
  • Pay-as-you-go research workflows with no subscription commitment
  • Developers integrating web-grounded QA into agent pipelines via MCP/x402

Not for

  • High-volume batch QA where per-request costs add up (consider direct Exa API subscription)
  • Use cases requiring real-time streaming chat (stream flag exists but endpoint is request-response oriented)
  • Queries requiring private or proprietary data not indexed on the public web

Quick start

# Using agentcash CLI
npx agentcash try https://stableenrich.dev

# Or via MCP fetch tool:
mcp__agentcash__fetch(
  url="https://stableenrich.dev/api/exa/answer",
  method="POST",
  body={"query": "What is the capital of France?"}
)

Example

Request

{
  "query": "What is the capital of France?"
}

Response

{
  "answer": "The capital of France is Paris. It has been the capital since the 10th century and is the country's largest city.",
  "citations": [
    {
      "id": "abc123",
      "url": "https://en.wikipedia.org/wiki/Paris",
      "text": "Paris is the capital and most populous city of France...",
      "title": "Paris - Wikipedia",
      "author": null,
      "publishedDate": null
    }
  ],
  "costDollars": {
    "total": 0.01
  }
}

Endpoint

Transporthttp
Protocolmpp
CurrencypathUSD

Quality

0.92/ 1.00

Full OpenAPI 3.1 schema with request/response definitions, clear pricing ($0.01/request in USD), detailed guidance documentation, and example request bodies are all present. The endpoint returned 405 on HEAD/GET probes (expected since it is POST-only), but the OpenAPI spec and pricing table confirm it is live. Minor deduction because no actual 402 challenge was captured for this specific endpoint.

Warnings

  • Probe returned 405 on HEAD and GET — endpoint is POST-only, which is expected and documented in the OpenAPI spec
  • No direct 402 challenge was captured for this endpoint; liveness inferred from OpenAPI spec and platform documentation

Citations

Provenance

Indexed frommpp_dev
Enriched2026-04-19 16:16:01Z · anthropic/claude-opus-4.6 · v2
First seen2026-04-18
Last seen2026-04-22

Agent access