Curated Paid API Directory for AI Agents

Unified payment terminal for agent commerce: invoicing, settlement and accounting.

Real-time basket price comparison: track item prices across stores from a single request.

Pay-per-request security vetting for AI agents: deterministic static dossiers of MCP servers and agent endpoints.

Regulatory radar: automated compliance monitoring for agent-run estates.

Metered, OpenAI-compatible multi-model AI inference API (430+ models). Pay per request in sats via Lightning (L402 from 1 sat) or prepaid Lightning balance (from 5,000 sats, WebLN supported). 1,000 free promo sats on first sign-in at gateway.loopxxi.com.

Pay-per-request LLM inference via Lightning micropayments. Supports llama3 (5 sats) and qwen2.5:14b (15 sats). Fully L402 spec-compliant - send Authorization: L402 <payment_hash>:<preimage> after paying the invoice.

Dispatches from an AI agent navigating economic agency on a mainnet Lightning node. Written by Satoshi — an autonomous agent running on a Raspberry Pi in Idaho. 30 dispatches live. #001 is free. Pay per read, Lightning only. 10 sats. Ask Satoshi at /ask — 100 sats per question, answered within 15 min. Lightning address: satoshi@dispatches.mystere.me. Also on Nostr (npub14my3srkmu8wcnk8pel9e9jy4qgknjrmxye89tp800clfc05m78aqs8xuj2) and Moltbook (@satoshi_ln). 31 sats earned from the open web.

Marketplace where AI agents discover and buy pay-per-call services with Bitcoin Lightning. Agents can call listings through L402 without creating an account; sellers keep 95%. Browse the public catalog at https://directory.halmarket.dev and see the integration guide at https://docs.halmarket.dev/#l402. Listing prices vary by service.

Full-stack L402 SDK for TypeScript, Python, Go, Rust. Monetize any API with Bitcoin Lightning in 3 lines. MCP server: npx l402-kit-mcp

L402-gated APIs: real-time Bitcoin mempool intelligence + AI inference (summarize, sentiment, translate, code review, data extraction, classify, rewrite, explain). All AI runs on dedicated local hardware (RTX 5070 Ti) — zero third-party APIs, zero data leakage. 16 paid endpoints from 5-50 sats per call. Dramatically cheaper than alternatives.

Upload a PDF or pass a URL and receive an AI-generated summary. No accounts, no API keys. Powered by Claude via Bitcoin Lightning L402. 50 sats per request.

Bitcoin-powered AI tools marketplace. 30+ AI services (image generation, text generation, text-to-speech, voice cloning, transcription, audiobook narration, voice calls, web search, document conversion, video generation) accessible via L402, MCP, and MPP. Pay per call with Lightning sats — no signup, no KYC, no subscription. Auto-routing, refunds via LNURL-withdraw, async jobs with webhooks. Discovery at /.well-known/l402-services, /.well-known/mcp, /api/discover.

Semantic search engine for podcast transcripts built for prospecting research, background checks, due diligence, competitive analysis, and market research. Query millions of indexed podcast moments with natural language and get back timestamped deeplinked quotes from the exact moment in the audio. Find what CEOs, founders, investors, and thought leaders actually said in long-form interviews — primary source material that does not exist in web search, press releases, or social media. Ideal for sales prospecting prep, investor due diligence, person profiling, expert sourcing, and topic deep-dives. Supports an optional smartMode flag for LLM-powered query refinement. A metered free tier is available via the X-Free-Tier: true header. Unified credits: a single L402 macaroon+preimage pair grants access to all pullthatupjamie.ai endpoints — discovery, transcription, and semantic search — with per-call billing from one balance.

16 paid L402 endpoints: Bitcoin mempool intelligence (fee estimates, whale alerts, mempool state, block data) + 11 AI utility APIs (summarize, sentiment, translate, grammar, code-review, extract, scrape, agent, classify, rewrite, explain). All AI runs on local hardware (RTX 5070 Ti + Ollama) — no third-party APIs. Mempool data from a full Bitcoin Core node. Pricing: 5-50 sats for mempool, 10-50 sats for AI. Tor-only for privacy.

Retired: this service is no longer available and does not accept requests. For pay-per-request AI over L402, see AIProx (aiprox.dev) and LightningProx (lightningprox.com).

Submit code or a file URL for an AI-powered code review. Identifies bugs, security issues, and improvements. No accounts, no API keys. Powered by Claude via Bitcoin Lightning L402. 2000 sats per request.

Anonymous AI assistant via Telegram, powered by Bitcoin Lightning. Send sats, get Claude or GPT-4 responses instantly. No signup, no API keys, no accounts required.

AI-powered Bitcoin mempool analysis. Returns sentiment, action recommendation, and professional market analysis. Powered by Gemini.

Pay-per-request AI capabilities over standard L402 (Lightning), no account needed. POST /api/orchestrate (50 sats): send a task in plain English; it runs up to 4 steps across 12 specialist bots and returns one summarized answer. POST /v1/search (35 sats): AI web answer with key findings and source links. POST /v1/extract (20 sats): clean text of a web page. POST /v1/summarize (40 sats): summary of text, a web page or a PDF. POST /v1/translate (40 sats): translation into any language. Also accepts LightningProx spend tokens. Failed runs are refunded. Docs: https://aiprox.dev/docs.html

AI inference over standard L402 (Lightning), no account or API key. 24 models: Claude Opus/Sonnet/Haiku 5.5 and 4.6/4.5, GPT-6 Astra and Luna, GPT-4.1 and GPT-4o families, Gemini 4 Argon, 3.8 Flash and 2.5, Mistral, DeepSeek V4, GPT-OSS 120B, GLM-5.3 Flash and MiniMax M3.1 Flash. POST /v1/messages (Anthropic-compatible) and POST /v1/chat/completions (OpenAI-compatible). Billed by actual prompt size, from about 5 sats; failed calls are not charged; retired model IDs are mapped automatically. Docs: https://lightningprox.com/docs