MPPtempoquality 0.62

Pay-per-call access to xAI Grok models for chat completions via Locus MPP on Tempo L2.

Price
per_call
Protocol
mpp
Verified
no
Endpoint
https://grok.mpp.paywithlocus.com/grok/chat

What it does

This endpoint provides chat completions powered by xAI's Grok family of large language models, accessed through the Locus Micropayment Protocol (MPP) with settlement on Tempo L2. It supports multiple model variants including grok-3, grok-3-mini, grok-4-0709, grok-4-1-fast-reasoning, and grok-code-fast-1. The request format follows the familiar OpenAI-compatible chat completions pattern with messages, temperature, top_p, max_tokens, stop sequences, frequency/presence penalties, and multi-completion support. Multimodal input (text + image URLs) is also supported in message content.

This chat endpoint is part of a broader Grok service suite hosted at grok.mpp.paywithlocus.com that also includes image generation, image editing, web search-augmented chat, X (Twitter) search-augmented chat, code execution-augmented chat, and text-to-speech — each at their own path. Pricing for the chat completions endpoint is model-dependent, estimated at roughly $0.001–$0.02 per call. Payment is handled via the MPP charge intent using pathUSD on Tempo L2, meaning each request is individually settled with no subscription or API key required.

Note that the probe did not capture a live 402 challenge on HEAD/GET for this specific endpoint; the endpoint is POST-only, so the 404 on HEAD/GET is expected behavior. The OpenAPI spec is well-defined with full request schemas and payment metadata. Documentation is referenced at docs.x.ai and beta.paywithlocus.com/mpp/grok.md, though the crawl did not retrieve those pages directly.

Capabilities

chat-completionsmultimodal-inputmulti-model-selectiontemperature-controlnucleus-samplingstop-sequencesfrequency-penaltypresence-penaltympp-paymenttempo-l2-settlementopenai-compatible-api

Use cases

  • —Agent-driven conversational AI where each query is paid individually without API keys
  • —Multimodal chat combining text and image inputs for visual question answering
  • —Rapid prototyping with multiple Grok model variants (mini, fast-reasoning, code-focused)
  • —Building chatbots or assistants that leverage xAI's latest Grok models
  • —Programmatic content generation with fine-grained sampling controls

Fit

Best for

  • —AI agents needing pay-per-call LLM access without subscription management
  • —Developers wanting OpenAI-compatible chat completions from xAI Grok models
  • —Applications requiring crypto-settled micropayments for each inference call

Not for

  • —High-volume batch processing where subscription-based pricing is more economical
  • —Use cases requiring streaming responses (no streaming documented for this endpoint)
  • —Users who need traditional API-key authentication instead of micropayment settlement

Quick start

curl -X POST https://grok.mpp.paywithlocus.com/grok/chat \
  -H "Content-Type: application/json" \
  -d '{
    "model": "grok-3",
    "messages": [
      {"role": "user", "content": "Explain quantum entanglement in one paragraph."}
    ],
    "max_tokens": 256
  }'

Example

Request

{
  "model": "grok-3",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful assistant."
    },
    {
      "role": "user",
      "content": "What is the capital of France?"
    }
  ],
  "max_tokens": 128,
  "temperature": 0.7
}

Endpoint

URLhttps://grok.mpp.paywithlocus.com/grok/chat
Transporthttp
Protocolmpp
CurrencypathUSD

Quality

0.62/ 1.00

The OpenAPI spec is detailed with full request schemas, model lists, and payment metadata for all seven endpoints. However, the probe did not capture a live 402 MPP challenge (endpoint is POST-only, probed with HEAD/GET), no response schema is documented, no example responses are available, and the external docs pages were not crawled. Pricing is approximate and model-dependent.

Warnings

  • —Probe returned 404 on HEAD/GET — endpoint is POST-only; live 402 challenge was not captured
  • —No response schema documented in the OpenAPI spec
  • —Pricing is described as 'model-dependent (~$0.001–$0.02)' without exact per-model breakdown
  • —External documentation at docs.x.ai and beta.paywithlocus.com/mpp/grok.md was not crawled to verify claims
  • —Streaming support is not documented — unclear if stream parameter is accepted

Citations

Provenance

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

Agent access

JSONhttps://clawmart.sh/api/listings/H3RqdZ