MPPtempoquality 0.55

Grok-4 chat completions augmented with live web search, paid per call via Locus MPP on Tempo L2.

Price
per_call
Protocol
mpp
Verified
no

What it does

This endpoint provides chat completions powered by xAI's Grok-4 family models with integrated web search capabilities. When you send a conversation to `/grok/web-search`, the model can query the live web to ground its responses in current information. You can constrain searches to specific domains (up to 5 allowed or 5 excluded) via the `search_parameters` field. Supported models include `grok-4-0709`, `grok-4-1-fast-reasoning`, and `grok-4-1-fast-non-reasoning`.

Payment is handled through the Locus Micropayment Protocol (MPP) settling on Tempo L2 in pathUSD. Pricing is dynamic, estimated at roughly $0.01–$0.50 per request depending on model choice and token usage. The request schema follows an OpenAI-compatible chat completions format with `model`, `messages`, optional `temperature` (0.0–2.0), and `max_tokens` parameters.

This endpoint is part of a broader Grok service suite hosted at `grok.mpp.paywithlocus.com` that also offers plain chat completions, X/Twitter search, code execution, image generation/editing, and text-to-speech — each at its own path. Documentation is available at https://docs.x.ai for the underlying xAI API and https://beta.paywithlocus.com/mpp/grok.md for the MPP-specific integration.

Capabilities

web-search-augmented-chatgrok-4-modelsopenai-compatible-chat-formatdomain-filteringdynamic-pricingtempo-l2-settlementper-call-payment

Use cases

  • Answering questions that require up-to-date web information
  • Research tasks where the LLM needs to cite current sources
  • Building agents that combine reasoning with live web lookups
  • Fact-checking or verifying claims against current web content
  • Generating summaries of recent events or news

Fit

Best for

  • AI agents needing grounded, web-current answers from Grok-4 models
  • Developers who want pay-per-call access without API key management
  • Applications requiring domain-scoped web search within LLM responses

Not for

  • High-volume batch processing where dynamic pricing may be unpredictable
  • Use cases requiring non-Grok-4 models (web search requires Grok-4 family)
  • Offline or latency-critical applications where web search adds delay

Quick start

curl -X POST https://grok.mpp.paywithlocus.com/grok/web-search \
  -H "Content-Type: application/json" \
  -d '{
    "model": "grok-4-0709",
    "messages": [
      {"role": "user", "content": "What are the latest developments in fusion energy?"}
    ],
    "max_tokens": 1024
  }'

Example

Request

{
  "model": "grok-4-0709",
  "messages": [
    {
      "role": "user",
      "content": "What is the current price of Bitcoin?"
    }
  ],
  "max_tokens": 512,
  "temperature": 0.7,
  "search_parameters": {
    "allowed_domains": [
      "coinmarketcap.com",
      "coingecko.com"
    ]
  }
}

Endpoint

Transporthttp
Protocolmpp
CurrencypathUSD

Quality

0.55/ 1.00

The OpenAPI spec provides a clear schema, model list, pricing guidance, and documentation links. However, the probe did not capture a live 402 challenge on HEAD/GET (the endpoint is POST-only, so 404 on HEAD/GET is expected), no example response is available, and all crawled pages returned 404 with no additional documentation content. Pricing is described as dynamic with a wide range.

Warnings

  • Probe returned 404 on HEAD and GET — endpoint is POST-only; liveness via MPP 402 challenge was not confirmed
  • Pricing is described as 'Dynamic (~$0.01–$0.50)' with no fixed amount; actual cost per call may vary significantly
  • No example response schema is provided in the OpenAPI spec
  • search_parameters is typed as string in the schema but described as an object with allowed_domains/excluded_domains — possible schema inaccuracy

Citations

Provenance

Indexed frommpp_dev
Enriched2026-04-19 17:22:56Z · anthropic/claude-opus-4.6 · v2
First seen2026-04-18
Last seen2026-04-22

Agent access