Perplexity Sonar chat with real-time web grounding, pay-per-call via Locus MPP on Tempo L2.
What it does
This endpoint exposes Perplexity's Sonar chat models through the Locus MPP (Micropayment Protocol), settling payments in pathUSD on Tempo L2. The `/perplexity/chat` path accepts OpenAI-compatible chat-completion requests with support for multiple Sonar model variants: `sonar`, `sonar-pro`, `sonar-reasoning-pro`, and `sonar-deep-research`. Each request can include web-search options (context size, search type, user location), domain/recency/language filters, structured JSON output via `response_format`, image return, related-question generation, and configurable reasoning effort for the reasoning-pro model. Pricing is model-dependent, roughly $0.005–$0.02 per call.
The same Locus MPP gateway also hosts sibling endpoints for standalone web search (`/perplexity/search` at ~$0.006/call), standard embeddings (`/perplexity/embed` at ~$0.001/call using `pplx-embed-v1-0.6b` or `pplx-embed-v1-4b`), and contextualized embeddings (`/perplexity/context-embed` at ~$0.001/call). All endpoints use the `charge` intent, meaning each call is a one-shot payment.
Note: The probe did not capture a live 402 challenge on HEAD/GET for the `/perplexity/chat` path (it returned 404), which is expected because this endpoint only accepts POST requests. The OpenAPI spec is well-defined with full request schemas, and the gateway's error messages point to valid documentation at `paywithlocus.com/skill.md` and `paywithlocus.com/llms.txt`.
Capabilities
Use cases
- —Agent-driven research queries with real-time web citations
- —Retrieval-augmented generation using web-grounded Sonar models
- —Generating embeddings for semantic search over documents
- —Answering time-sensitive questions with recency-filtered web results
- —Structured data extraction from the web via JSON schema output
Fit
Best for
- —AI agents needing web-grounded answers without a Perplexity API key
- —Pay-per-call access to Perplexity Sonar without subscriptions
- —Applications requiring cited, up-to-date web information in chat responses
- —Developers wanting crypto-settled micropayments for LLM inference
Not for
- —High-volume batch processing where subscription pricing is cheaper
- —Use cases requiring non-Perplexity models (GPT-4, Claude, etc.)
- —Applications that cannot settle payments on Tempo L2 / pathUSD
Quick start
curl -X POST https://perplexity.mpp.paywithlocus.com/perplexity/chat \
-H "Content-Type: application/json" \
-H "Authorization: Bearer <MPP_TOKEN>" \
-d '{
"model": "sonar",
"messages": [{"role": "user", "content": "What happened in tech news today?"}]
}'Example
Request
{
"model": "sonar-pro",
"messages": [
{
"role": "system",
"content": "Be concise and cite sources."
},
{
"role": "user",
"content": "What are the latest developments in fusion energy?"
}
],
"max_tokens": 1024,
"temperature": 0.7,
"web_search_options": "{\"search_context_size\": \"medium\", \"search_type\": \"auto\"}",
"search_recency_filter": "week",
"return_related_questions": true
}Endpoint
Quality
The OpenAPI spec is detailed with full request schemas, model options, and approximate pricing. However, the probe did not capture a live 402 challenge (POST was not attempted), no response schema or example response is documented, and crawled pages all returned 404 (expected for non-endpoint paths). Pricing is approximate and amount fields are null in the spec.
Warnings
- —Probe returned 404 on HEAD/GET; endpoint is POST-only so liveness via 402 challenge was not confirmed
- —Pricing amounts are listed as null in x-payment-info; only human-readable approximations are given in description fields
- —No response schema or example response is provided in the OpenAPI spec
- —Documentation links (paywithlocus.com/skill.md, paywithlocus.com/llms.txt) were not crawled to verify availability
Citations
- —Supports models: sonar, sonar-pro, sonar-reasoning-pro, sonar-deep-researchhttps://perplexity.mpp.paywithlocus.com
- —Chat pricing is model-dependent, roughly $0.005–$0.02 per callhttps://perplexity.mpp.paywithlocus.com
- —Web search endpoint costs ~$0.006 per callhttps://perplexity.mpp.paywithlocus.com
- —Embedding endpoints cost ~$0.001 per callhttps://perplexity.mpp.paywithlocus.com
- —Settlement via Tempo L2 using pathUSD (charge intent)https://perplexity.mpp.paywithlocus.com
- —Documentation available at paywithlocus.com/skill.md and paywithlocus.com/llms.txthttps://perplexity.mpp.paywithlocus.com
- —Perplexity API reference at docs.perplexity.aihttps://perplexity.mpp.paywithlocus.com