Sats4AI - Bitcoin-Powered AI Tools

Send an AI voice agent to make a two-way phone call. The agent follows your task instructions, navigates IVR menus, has a real conversation, and returns a full transcript with structured analysis. Automatically retries on voicemail/no-answer (up to 3 attempts, included in payment). May return state='pending_confirm' with clarification questions if the task is missing critical info (e.g., booking without date/time) — call confirm_ai_call to provide answers and proceed. On task failure, get_job_result returns the full transcript so you can adjust and retry with a new payment. Async — returns call_id or session_id, poll with check_job_status(jobType='ai-call'). Requires create_payment with toolName='ai_call', phoneNumber, and durationMinutes. Dynamic pricing: telephony cost + AI agent cost per minute (~150-250 sats for a 3-min US call). Languages: en-US, en-GB, es-ES, fr-FR, de-DE, ja-JP, zh-CN, multi (voice auto-selected per language).

Analyze and describe image content using Qwen VL. Returns text analysis. Requires create_payment with toolName='analyze_image'. Default model: id=5 (21 sats). Send image as base64.

Animate a still image into video with text guidance using Grok Imagine Video. Async — returns requestId, poll with check_job_status. Requires create_payment with toolName='animate_image'. Default model: id=24 (100 sats/sec). Duration 3-15 seconds. Requires base64 image.

Check the status of an async job (generate_video, animate_image, generate_3d_model, transcribe_audio, epub_to_audiobook, ai_call). Returns status: queued/processing/completed/failed. For epub-audiobook, also returns progress percentage and chapter-level progress.

Check if a Lightning payment has been received. Returns paid/pending status.

Clone a voice from an audio sample using Speech-02 Turbo or Speech-02 HD. Returns a voice_id that can be used with the text_to_speech tool for text-to-speech in the cloned voice. Requires create_payment with toolName='clone_voice'. 7500 sats. Provide a public URL to an audio file (MP3, WAV). Processing takes 1-5 minutes.

Confirm an AI call after reviewing push-back questions, optionally providing answers to missing info. Required when ai_call returns state='pending_confirm'. Uses the original payment — no new payment needed. Returns call_id for polling with check_job_status(jobType='ai-call').

Convert a file between formats. Supports 200+ formats: documents (PDF, DOCX, XLSX), images (PNG, JPG, WEBP, SVG), audio (MP3, WAV, FLAC), video (MP4, AVI, MOV). Returns download URL. 100 sats. Requires create_payment with toolName='convert_file'. Provide file as base64 or URL.

Convert HTML or Markdown to a PDF document. Returns download URL. Requires create_payment with toolName='convert_html_to_pdf'. 50 sats. Great for generating invoices, reports, or formatted documents from code.

Create a Lightning invoice to pay for one AI service call. Each payment covers exactly one output. Call list_models first to see available models. modelId is optional — omit it to use the default (best) model. For generate_text, you must provide the exact prompt now (price is calculated from character count). For send_sms, place_call, and ai_call, phoneNumber is required at payment time. For ai_call, also provide durationMinutes (1-10, default 3). For generate_video, you must provide duration, mode, and generate_audio at payment time (price depends on these). For animate_image, provide duration (100 sats/sec). For edit_image, provide resolution (1K=200, 2K=300, 4K=450 sats).

Edit an image using Nano Banana 2. Describe the changes you want in the prompt. Returns base64 edited image. Requires create_payment with toolName='edit_image' and resolution param. Resolution-tiered pricing: 1K=200 sats, 2K=300 sats, 4K=450 sats.

Convert an EPUB book into a full AI-narrated audiobook. Async — returns jobId, poll with check_job_status(jobType='epub-audiobook'). Requires create_payment with toolName='epub_to_audiobook'. Dynamic pricing: ~1 sat per 33 characters, minimum 500 sats. Send EPUB as base64. Returns zip of chapter MP3s. Processing may take 5-60+ minutes depending on book length.

Extract text from PDFs and images as clean Markdown using Mistral OCR. Returns Markdown text and page count. Requires create_payment with toolName='extract_document' and quantity=pageCount for multi-page PDFs. Default model: id=27 (10 sats/page). Send document as base64.

Extract structured data from receipts, invoices, and financial documents using Mistral OCR + Kimi K2.5. Returns JSON with merchant, date, line items, totals, tax, currency, and category. 50 sats/page (use quantity param in create_payment for multi-page PDFs). Requires create_payment with toolName='extract_receipt'. Send document as base64.

Convert a photo into a downloadable 3D GLB model using ByteDance Seed3D. Async — returns requestId, poll with check_job_status. Requires create_payment with toolName='generate_3d_model'. Default model: id=17 (350 sats). Send image as base64 PNG/JPEG/WEBP. May take up to 5 minutes.

Generate an image from a text prompt. Returns base64 PNG. Requires create_payment first with toolName='generate_image'. Models: id=2 xAI Grok Imagine (fast, 100 sats), id=8 ByteDance Seedream 4 (detailed, 150 sats), id=16 Google Nano Banana 2 (premium, 200 sats, default best). Omit modelId for default. Supports optional imageBase64 for img2img.

Generate a song with AI vocals and instrumentation using Music-1.5. Returns MP3 URL. Requires create_payment with toolName='generate_music'. Default model: id=23 (100 sats). Provide a style prompt and optional lyrics.

Generate text using AI language models. Returns text response. Requires create_payment with toolName='generate_text' and the exact prompt. Dynamic pricing (pure cost per character, no minimum): best Kimi K2.5 (id=6, 100 chars/sat, 262K context, vision support), standard GPT-OSS-120B (id=1, 333 chars/sat), budget Qwen3-32B (id=26, 1000 chars/sat, 119 languages). Omit modelId for default (best). Supports fileContext for document Q&A and imageBase64 for vision (best model only).

Generate a video from a text prompt using Kling v3. Async — returns requestId, poll with check_job_status. Requires create_payment with toolName='generate_video' and duration, mode, generate_audio params. Pricing: standard 300-400 sats/sec, pro 450-550 sats/sec (audio adds 100 sats/sec). Duration 3-15 seconds.

Get the final result of a completed async job. Call only after check_job_status returns completed.

Get pricing for a specific model by ID. No payment required.

List available AI models with IDs, tiers, descriptions, and pricing. Call this before create_payment to discover models. Each model has a stable numeric ID. Models marked isDefault=true are used when you omit modelId. Filter by category to see models for a specific tool.

List all planned services with current vote counts. No payment required. Use this to discover what services are coming and which slugs to use with vote_on_service. Services are sorted by vote count (most-voted first).

Merge multiple PDF files into one. Returns download URL. Requires create_payment with toolName='merge_pdfs'. 100 sats. Send PDFs as base64-encoded strings in an array (minimum 2 files).

Place an automated phone call. Delivers a spoken message (TTS) or plays an audio file. Requires create_payment with toolName='place_call' and phoneNumber. Provide message (TTS, max 500 chars) OR audioUrl (public URL to audio file).

Request a refund for a failed service. Only works if the payment is marked as refund-eligible (check_payment_status shows refundAllowed=true). Provide the paymentId and a Lightning address or bolt11 invoice where the refund should be sent. Refunds are reviewed manually and processed within 48 hours. No payment required to submit a refund request.

Send an email to any address. No account or SMTP setup required. Flat rate: 200 sats. Requires create_payment with toolName='send_email'. Plain text only, max 10,000 chars body, 200 chars subject. A footer is appended automatically.

Send an SMS message to any phone number worldwide. Requires create_payment with toolName='send_sms' and phoneNumber+message at payment time. The phoneNumber and message must match those used in create_payment.

Convert text to natural speech audio using Speech-02-HD. Returns MP3 URL. Requires create_payment with toolName='text_to_speech'. Default model: id=3 (300 sats). High-fidelity synthesis with natural cadence and emotional expression. Multiple voice options. Speed 0.5-2.0.

Transcribe audio to text using Mistral Transcription. Async — returns requestId, poll with check_job_status(jobType='transcription'), then get_job_result. Requires create_payment with toolName='transcribe_audio'. Default model: id=10 (10 sats/min). Supports 13 languages: en, zh, hi, es, ar, fr, pt, ru, de, ja, ko, it, nl. Send audio as base64. Max file size: 512 MB. Up to 3 hours per file.

Translate text across 119 languages using Qwen3-32B. Privacy-preserving: no data stored. Pricing: 1,000 characters per sat, pure cost per character (no minimum). Requires create_payment with toolName='translate_text' and prompt (the text to translate). Supports automatic source language detection. Popular languages: English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Chinese, Japanese, Korean, Arabic, Hindi, Turkish, Polish, Swedish, Norwegian, Danish, Finnish, Czech, Romanian, Hungarian, Ukrainian, Greek, Thai, Vietnamese, Indonesian, Malay, Hebrew, Farsi, Bengali, Tamil, Urdu, Swahili, Filipino, Georgian, Armenian, and many more.

Vote for a planned service to be built next. 1 sat per vote. Call list_planned_services first to see available slugs and current vote counts. Requires create_payment with toolName='vote_on_service'. Highest-voted services get built first.