Sats4AI - Bitcoin-Powered AI Tools

When your task hits a wall that requires a human — booking, negotiating, navigating IVR menus, getting information from a business — send an AI voice agent to handle the call. The agent follows your instructions, has a real two-way conversation, and returns a full transcript with structured analysis. One redial, 2 minutes later, if nobody answers, a voicemail answers, the line is busy or the call drops within 30 seconds. No redial after the person speaks and hangs up, asks to stop or opts out, or after the agent leaves a voicemail message. A person who asks not to be called again is added to our do-not-contact list. The redial runs on the deposit's unused minutes (none left: no redial), and a connected redial bills at least 1 minute. May return state='pending_confirm' with clarification questions if critical info is missing — call confirm_ai_call to proceed. Async — poll with check_job_status(jobType='ai-call'). Priced in sats per destination and duration — create_payment returns the exact amount before you pay. Languages: en-US, en-GB, es-ES, fr-FR, de-DE, ja-JP, zh-CN, multi. Pay with Bitcoin Lightning — no telecom account, no API key, no subscription. When NOT to use: not when you want to drive the conversation with your own LLM (use open_voice_bridge — you keep the brain, we provide PSTN/STT/TTS primitives). Not for one-shot TTS broadcasts or IVR delivery (use place_call). Not for SMS (use send_sms). Requires create_payment with toolName='ai_call', phoneNumber, and durationMinutes. The agent identifies itself as a virtual assistant calling for a client, and answers questions about the service if the recipient asks. At the end of a normal call it asks the recipient for permission to share the recording; without a clear yes, the recording is deleted and never returned. No impersonation, scams, threats, harassment, credential theft, unsolicited bulk messages, or contact after an opt-out. Answering a call does not establish consent. We screen requests after payment. Held messages and calls are not sent or placed; payment is retained during review, not refunded immediately. Confirmed violations are not refunded. Unresolved communication holds become eligible for a refund after 24 hours; refund eligibility is processed by scheduled maintenance. Human review costs 1 sat at https://sats4ai.com/appeal. Recent message text and extracts of recording transcripts support recipient-scoped abuse checks for 24 hours. These context records are deleted hourly, so storage can last up to 25 hours. Original uploaded-call transcripts and raw held-request or appeal evidence enter cleanup after 30 days. We also keep a de-identified record of screened communications to measure and improve fraud detection. Identifiers in it — contact addresses, numbers and names that follow a greeting — are replaced with one-way tokens, and it holds no destination number and no account, because the service has none. De-identification is not anonymization: text a sender wrote can still identify someone. Do not submit personal information the service does not need. Providers keep records under their own policies.

Analyze and describe image content, answer visual questions, extract information from screenshots or photos. Uses Qwen VL — multimodal vision-language model with strong OCR, chart reading, and spatial reasoning. 21 sats per image. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='analyze_image'.

Animate a still image into cinematic video with MiniMax H3 — provide a first frame and a prompt to direct the motion. Native audio. Async — returns requestId, poll with check_job_status. 768p or 2K (2K is upscaled from a 768p render), duration 5-15 seconds, priced per second by resolution. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='animate_image'.

Wait for an async job to finish and return its result in ONE call — no manual polling loop. Pass the requestId + jobType from an async tool (generate_video, animate_image, generate_3d_model, transcribe_audio, epub_to_audiobook, ai_call). If your MCP client opened the request with 'Accept: text/event-stream', this streams live progress (notifications/progress) while it waits, then returns the final result; otherwise it does a single status check and returns immediately (call again until status='COMPLETED'). For long jobs it waits up to ~4 minutes per call, then returns status='IN_PROGRESS' with timed_out=true — call again with the same requestId to keep waiting. Equivalent to check_job_status + get_job_result combined. Free; no payment.

Turn airline boarding passes (PDF or screenshot) into Google Wallet passes with the fields a plain scan drops — boarding zone, gate, boarding time — for a whole family in one go. Returns one save link per passenger plus a single 'add all' link. The original barcode is preserved byte-for-byte so it scans at the gate. Priced 100 sats PER PASS: create_payment with toolName='boardingpass_wallet' and quantity=<number of passes> (if you under-count, the call returns the exact price to re-pay). Image/screenshot files that turn out to carry no extra fields are refunded. Privacy: files are processed in memory and never stored; the save link expires in 30 minutes, so nothing is retained. Pay with Bitcoin Lightning — no account, no card.

Poll the status of an async job. Use this after calling any async tool (generate_video, animate_image, generate_3d_model, transcribe_audio, epub_to_audiobook, ai_call) that returns a requestId. Returns JSON: { status: 'IN_PROGRESS' | 'COMPLETED' | 'FAILED', requestId, jobType }. For epub-audiobook, also includes progress (0-100) and chapterProgress array. Poll every 5-10 seconds. When status is 'COMPLETED', call get_job_result to retrieve the output. When status is 'FAILED', the response includes an error message — do not retry automatically. This tool is free and does not require payment. Do NOT use for synchronous tools (generate_image, generate_text, etc.) — those return results immediately.

Check whether a Lightning invoice has been paid. Returns JSON: { paid, serviceUsed, refundAllowed, readyToUse, next, state } — readyToUse=true means paid, not yet consumed, and no refund queued; next says what to do. state is unpaid, paid, delivered, held, refund-open, claimed or expired. held (with its appeal_url) is present while the payment is held for review. deliveryStatus and deliveryDetail are present for SMS and fax: a send that succeeded means accepted, not delivered, so poll for the carrier outcome. Call after create_payment to verify the user has paid before calling the target tool. Invoices expire after 10 minutes — if expired, create a new payment. Most MCP clients with a connected wallet pay instantly, so a single check is usually sufficient. This tool is free and does not require payment.

Clone a voice you own or have permission to use, from a single audio sample. Returns a reusable voice_id for text_to_speech. Requires consent=true; cloning a voice to impersonate someone is prohibited. High-fidelity reproduction capturing tone, cadence, and accent. SAMPLE REQUIREMENTS: MP3, M4A or WAV, 10 seconds to 5 minutes, 20 MB max, one speaker, no background music — a sample outside these limits is rejected upstream and the clone fails. Turbo (faster) or HD (higher quality) modes. 7,500 sats per clone. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='clone_voice'.

Colorize black-and-white or grayscale photos. DDColor (dual-decoder, ICCV 2023) — vivid, natural colorization. Impossible for text/vision LLMs. 5 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='colorize_image'.

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 files between 200+ formats: documents (PDF, DOCX, XLSX, HTML→PDF), images (PNG, JPG, WEBP, SVG), PDF first page to image (PDF→JPG, PDF→PNG), audio (MP3, WAV, FLAC), video (MP4, AVI, MOV). Industrial-grade conversion engine — preserves formatting and quality. Returns download URL. 100 sats. Pay per request with Bitcoin Lightning — no API key, no account, no subscription needed. Requires create_payment with toolName='convert_file'.

Convert HTML or Markdown to a pixel-perfect PDF. Returns JSON: { url } — a temporary download URL (valid ~1 hour). Great for generating invoices, reports, receipts, or formatted documents programmatically. Supports full HTML/CSS including tables, images (base64 or URL), and inline styles. For Markdown input, set format='markdown'. 50 sats per conversion. Use convert_file instead for converting existing files between formats (e.g., DOCX→PDF). Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='convert_html_to_pdf'.

Pay for 2 to 10 calls with ONE Lightning invoice, up to 2,000 sats in all. Each item is { toolName, ...the same inputs create_payment takes for that tool }. The server checks and prices EVERY item first: one bad item refuses the whole cart, with no invoice, and the error names the item by its index. Returns JSON: { cartId, lightningInvoice, totalSats, expiresAt, items: [{ index, toolName, paymentId, amountSats }] }. Pay the ONE invoice. Then call each item's tool with that item's own paymentId, exactly as you would after create_payment. KEEP the paymentIds from this reply: each is the only way to use that item or to be refunded for it, and nothing can show them to you again. check_payment_status(cartId) shows the state of each item (it never repeats the paymentIds). An item you do not use is refunded for its own price after about an hour. Only one-shot tools can be in a cart (text, translation, image tools, document tools, text_to_speech); for calls, SMS, fax, email, video, music and other long jobs use create_payment.

Create a Lightning invoice to pay for one AI service call. Returns JSON: { paymentId, lightningInvoice (BOLT11), amountSats, expiresAt }. Each payment covers exactly one tool call — call this once per operation. Typical flow: list_models → create_payment → check_payment_status → call tool. The invoice expires in 10 minutes. Call list_models first to discover modelId values. modelId is optional — omit it to use the default (best) model. Some tools require extra params at payment time because pricing depends on them: generate_text requires prompt (price = f(char count)); translate_text requires prompt (the text to translate) and takes targetLanguage (it picks the engine and so the price); text_to_speech requires text (price = f(char count) by tier); transcribe_audio / transcribe_translate take durationMinutes (10 sats/min — declare your audio length, default 1); send_sms, place_call, ai_call require phoneNumber; generate_video and animate_image require duration, and take an optional resolution (250-400 sats/sec by resolution — quote with the SAME duration and resolution you will execute with); edit_image is a flat 200 sats per edit (resolution is optional and does not change the price); epub_to_audiobook requires characterCount (total text characters in the book — price is per-character by voice tier, minimum 500 sats). If required params are missing, the response includes an error with the missing field names.

Recover detail from camera-shake and accidental motion blur. NAFNet (ECCV 2022, SOTA on GoPro/SIDD benchmarks). Best for: handheld shake, bumped camera, whole-frame uniform blur. NOT effective for: intentional panning blur, bokeh/depth-of-field, or artistic motion effects. Also supports denoising (grainy/noisy photos). 110 sats per image (~2 min processing), pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='deblur_image'.

Classify image safety (normal / suggestive / explicit). Falcons.ai NSFW detection — 100x cheaper and faster than asking an LLM. Returns classification label and boolean is_nsfw flag. Essential for content moderation pipelines. 2 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='detect_nsfw'.

Detect and locate objects in an image by name. Grounding DINO (open-set detector, ECCV 2024) — describe what to find in natural language, get bounding box coordinates and confidence scores. Structured pixel data agents can't get from vision LLMs. 5 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='detect_objects'.

Send a document out for legally binding e-signature and have the signed copy emailed back. Upload the PDF, name up to 3 signers, and give the email where the signed copy should land. Flat 1000 sats per document. The request is fulfilled through an established e-sign provider. The signed copy arrives once every signer has signed, so expect their turnaround — hours to days — not seconds. The API reply acknowledges the request; there is nothing to poll. Privacy: the PDF is processed in memory and never stored. Pay with Bitcoin Lightning — no account, no card. Requires create_payment with toolName='e_signature'.

Edit an image with natural language instructions. Uses MAI-Image-2.6, ranked #3 on the Artificial Analysis image-editing leaderboard as of September 2026 — understands context, handles object addition/removal, style transfer, and inpainting. Returns JSON with image URL. 200 sats per edit, one flat price. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='edit_image'.

Hang up a Voice Bridge call, finalize billing, and return a LNURL-withdraw refund link for unused deposit time. Also returns the final transcript for convenience.

Convert books (EPUB/PDF/TXT) to full audiobooks with automatic chapter detection, multi-voice narration, and optional translation to any language before narration. 3 voice tiers: OmniVoice Global (602+ langs, ~106 chars/sat), Inworld Premium (#1 ranked TTS ELO 1217, ~16 chars/sat), Minimax Studio (voice cloning from reference clip, ~5 chars/sat). Min 500 sats. Async — returns jobId, poll until completed (5-60+ min). Single payment, full outcome — no multi-step orchestration required. Pay with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='epub_to_audiobook'.

Estimate per-pixel depth from a single photo. Depth Anything V2 (NeurIPS 2024, 3.8M+ runs) returns BOTH the raw greyscale depth map — what a ControlNet, AR or 3D reconstruction pipeline consumes — and a colour visualization for humans. Structured geometry a vision LLM cannot produce. 5 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='estimate_depth'.

Extract text from PDFs and images as clean Markdown. Uses Mistral OCR — handles complex layouts, tables, handwriting, multi-column documents, and mathematical notation. Preserves document hierarchy in structured Markdown. 10 sats/page. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='extract_document' and quantity=pageCount for multi-page PDFs.

Extract structured data from receipts, invoices, and financial documents. Uses a dual-model pipeline (Mistral OCR + Kimi K3) for high-accuracy extraction. Returns JSON with merchant, date, line items, totals, tax, currency, and expense category. Handles crumpled receipts, faded text, and multi-page invoices. 25 sats/page. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='extract_receipt'.

Generate a textured 3D GLB model from EITHER a photo OR a text prompt (provide exactly one, not both). Uses Tencent Hunyuan3D — high-fidelity geometry and PBR materials. Async — returns requestId, poll with check_job_status. 1600 sats per model. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='generate_3d_model'.

Generate an image from a text prompt. Returns JSON with image URL. Models: Grok Imagine 2 (fast creative generation, 100 sats), Seedream 5 Pro (most permissive content policy, 150 sats at 1K / 300 at 2K), GPT Image 2.6 (ranked #1 on the Artificial Analysis text-to-image leaderboard as of November 2026 at its maximum quality setting; this API uses the automatic setting; 1200 sats, default). Supports img2img with optional base64 input. Optional aspectRatio (default 1:1) works on every model tier and does not change the price. Stable endpoints — models upgrade automatically as SOTA evolves. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='generate_image'.

Generate a full song with AI vocals from a style prompt plus lyrics — or an INSTRUMENTAL bed / underscore (background music under video narration, intros, outros) by sending ONLY structure tags and no words. You choose the length: 15-300 seconds at 5 sats/second (75-1500 sats). Specify BPM, key, genre, mood in prompt. Returns a 44.1 kHz 16-bit stereo WAV URL. Structure tags must each be on their OWN line — words on a tag line are DROPPED by the model, which silently shortens the song. Don't want to write lyrics? Set lyricsAddon:true (+25 sats) and optionally lyricsAbout to say what they should be about. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='generate_music' AND the same duration (the price is bound to it).

Generate text using frontier AI language models. Pure per-character pricing (no minimum): Kimi K3 (best, ~10 chars/sat, 1M context, default), GPT-OSS-120B (standard, ~1000 chars/sat, 119 languages, best value). Rates are BTC-pegged and re-quoted hourly, so treat them as approximate — the 402 challenge is the authoritative price. Supports document Q&A via fileContext. For images, use the analyze_image tool. Stable endpoints — models upgrade automatically. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='generate_text' and the exact prompt.

Generate cinematic video from a text prompt. Uses MiniMax H3 — top 2 on the Artificial Analysis text-to-video leaderboard — with native audio. Async — returns requestId, poll with check_job_status. 768p or 2K (2K is upscaled from a 768p render), 5-15 seconds, priced per second by resolution (native audio free). Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='generate_video' and duration, resolution params.

Get an exact sat cost quote for a service BEFORE creating a payment. Useful for budget-aware agents to price-check before committing. No payment required, no side effects. Pass service=text-to-speech&chars=1500, service=translate&chars=800, service=transcribe-audio&minutes=5, etc. Returns { amount_sats, breakdown, currency }. Omit params to see the full catalog of supported services.

Get the machine-readable catalog of all error codes this API can return (e.g. TIMEOUT, CONTENT_FILTERED, RATE_LIMITED, L402_REFUND_ISSUED, L402_AUTO_ROUTED). Agents should branch on error_code rather than parsing free-text messages. No payment required.

Retrieve the final output of a completed async job. Call ONLY after check_job_status returns status='COMPLETED' — calling on a job that has not completed returns an error. Returns JSON whose shape depends on jobType: video/video-image → { videoUrl, duration }; image-3d → { modelUrl } (GLB format); transcription → { text, language, segments, srt, vtt } (a result too large for one reply comes back without segments, srt and vtt, with resultTruncated, omitted and fullResultUrl: a free link to the whole result); epub-audiobook → { url, annexeUrl, totalChapters, completedChapters } (annexeUrl is a companion PDF of equations, tables, addresses, keys and hashes the narration replaces with a spoken marker; null when the book had none); ai-call → { transcript, duration, summary }. A video or 3D job whose payment was refunded returns error_code RESULT_NOT_AVAILABLE, unless you pass its paymentId: the payer then gets the pending refund (error_code L402_REFUND_ISSUED). An audiobook whose payment was refunded also returns RESULT_NOT_AVAILABLE, with no link: claim the refund at /recover with the payment ID. A web 3D purchase not yet collected on its own page also returns RESULT_NOT_AVAILABLE. All URLs are temporary (valid ~1 hour) — download immediately. This tool is free and does not require payment. Do NOT use for synchronous tools — those return results directly.

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

Discover available AI models with numeric IDs, tier labels, capabilities, and per-call pricing in sats. Call this before create_payment to find the right modelId for your task. Returns JSON array: [{ id, name, tier, description, price, isDefault, category }]. Models marked isDefault=true are used when you omit modelId from create_payment. Filter by category to narrow results to a specific tool. This tool is free, requires no payment, and is idempotent — safe to call repeatedly.

List all planned services with current vote counts. Returns JSON array: [{ slug, name, description, votes }], sorted by votes descending. No payment required — this is a free discovery tool. Use the slug values with vote_on_service to cast votes. This tool is idempotent and safe to call repeatedly.

Merge multiple PDF files into a single document. Preserves bookmarks, links, and formatting. Returns JSON: { url } — a temporary download URL (valid ~1 hour). Minimum 2 files, no maximum. Files are concatenated in array order. 100 sats per merge regardless of file count. Use convert_file instead if you need format conversion (e.g., DOCX→PDF). Pay per request with Bitcoin Lightning — no API key, no account needed. Requires create_payment with toolName='merge_pdfs'.

Ask a frontier AI a question in your OWN language and get the answer back in that same language, in ONE payment. We translate the question to English, answer it with a frontier model, then translate the answer back. Two engines behind one endpoint: 119 mainstream languages use the LLM translate tier; anything those models do not serve falls through to MADLAD-400 on our own GPU (452 languages, 251 of which ChatGPT, Claude and Gemini do not support at all). So a Wolaytta, Tiv or Q'eqchi' speaker gets frontier reasoning in their own language, which no other API offers. Rare-language quality is uneven and published per language - GET https://sats4ai.com/api/languages. Priced from question length plus a fixed 2000-character answer allowance at the chosen tier's rate; the rare-language path costs more (GPU both ways). Requires create_payment with toolName='multilingual_ask' AND the `language` you intend to use - the quote depends on it.

Open a phone call you drive one turn at a time: we dial, transcribe what the other side says, and speak whatever text you send. You call voice_bridge_say to talk, poll_voice_bridge to read transcripts, end_voice_bridge to hang up. ⚠ Turn-based, NOT full-duplex conversation. After the callee stops talking, about 2 seconds of ours pass before your reply starts to play (we recognise the end of their turn, then speech starts under a second after you send the text), on top of your own model's thinking time: slower than the ~0.8s a natural back-and-forth needs. The callee cannot interrupt you while you speak. Pass filler (a short phrase such as 'One moment, please.') to cover a slow reply: we say it once if you have not started answering about 3.5 seconds after they stop. The loop: poll_voice_bridge with waitMs (e.g. 20000) returns when the callee has finished talking; answer with a short voice_bridge_say (one or two sentences); repeat. Good for: leaving a spoken message, navigating a voice-driven IVR (keypad tones are not supported yet), reading something out, a slow exchange where a pause is fine. For a real conversation with a human, use ai_call — we run the agent at conversational speed and return the transcript. Unused deposit time is refunded automatically when the call ends — including when the callee hangs up. Pass refundAddress (a Lightning address your user gave you; never invent one) and the remainder is SENT there with no further action; without it the refund is held as a claimable LNURL-withdraw link returned by end_voice_bridge. Use this when the content of the call must never leave your side, and the pace can tolerate a pause between turns. When NOT to use: not for fully-managed agent-style calls where we handle the brain (use ai_call). Not for one-shot TTS broadcasts or IVR playback (use place_call). Not when live transcript polling adds no value — the per-turn overhead isn't worth it. Privacy: transcripts held in memory only, garbage-collected 30 minutes after the call ends; call audio is never persisted. Pay with Bitcoin Lightning — no telecom account, no signup. Requires create_payment with toolName='voice_bridge_open', phoneNumber, durationMinutes. Deposit is priced per destination off the carrier rate sheet and is BTC-pegged, so call create_payment for the exact figure; premium-rate and satellite ranges are refused.

Bridge the digital-physical gap — place an automated phone call to deliver a spoken message or play audio to any number. Useful when your task requires notifying a human, delivering alerts, or reaching someone who isn't online. Pay with Bitcoin Lightning — no telecom account, no KYC, no subscription. Requires create_payment with toolName='place_call' and phoneNumber. The message plays first. Then a short spoken notice names the service and how to stop these calls: call the number back, or go to sats4ai.com/c. The notice fits the prepaid minutes. No impersonation, scams, threats, harassment, credential theft, unsolicited bulk messages, or contact after an opt-out. Answering a call does not establish consent. We screen requests after payment. Held messages and calls are not sent or placed; payment is retained during review, not refunded immediately. Confirmed violations are not refunded. Unresolved communication holds become eligible for a refund after 24 hours; refund eligibility is processed by scheduled maintenance. Human review costs 1 sat at https://sats4ai.com/appeal. Recent message text and extracts of recording transcripts support recipient-scoped abuse checks for 24 hours. These context records are deleted hourly, so storage can last up to 25 hours. Original uploaded-call transcripts and raw held-request or appeal evidence enter cleanup after 30 days. We also keep a de-identified record of screened communications to measure and improve fraud detection. Identifiers in it — contact addresses, numbers and names that follow a greeting — are replaced with one-way tokens, and it holds no destination number and no account, because the service has none. De-identification is not anonymization: text a sender wrote can still identify someone. Do not submit personal information the service does not need. Providers keep records under their own policies.

Fetch new transcript events from an open Voice Bridge call since the last cursor. Returns partial + final transcripts + system events. Pass waitMs (e.g. 20000) to hold the call until the callee has FINISHED TALKING (a final transcript, then 0.7 s of silence) or the call changes state, then get everything at once: no polling timer, and you do not answer in the middle of their sentence. Without waitMs, poll in a loop (~500ms-1s) and decide yourself when they are done. No additional payment.

When you're expecting a fax back — bank confirmation, court filing, signed document — open a 24h receive window at our shared number +1 320 299 1523. Matched by caller ID (last 10 digits of the sender), delivered to your email as soon as it arrives. Optional OCR add-on (+200 sats) returns a searchable text file alongside the PDF — useful for feeding the content to an agent or archiving. Optional callback_url POSTs an HMAC-signed webhook on delivery so your agent doesn't have to poll. No refund if no fax arrives within the window (prevents subscription squatting). If OCR fails, an LNURL-withdraw for 200 sats is included in the delivery email for partial refund. Pay with Bitcoin Lightning — no dedicated fax number rental, no monthly subscription, no account.

Remove background from any image, returning transparent PNG. Uses BiRefNet (state-of-the-art, Papers with Code — Sm 0.901 on DIS5K). Handles hair, fur, glass, transparency, and complex edges. Stable endpoint — model upgrades automatically as SOTA evolves. 44 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='remove_background'.

Remove unwanted objects from images by describing what to remove — no mask needed. Combines Grounding DINO detection (ECCV 2024) with Bria Eraser inpainting. Just say 'person', 'car', or 'watermark' and the object is erased and filled convincingly. 130 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='remove_object'.

Render a typographic card — headline, versus (two columns), list, stat (one big number) or quote — as a PNG or JPEG at 1920x1080 or 2560x1440. Deterministic layout, no AI model: the text you send is exactly what appears, which no image generator can promise. Built for video title cards and slides, thumbnails, quote cards and social posts. Lists and comparisons REVEAL one beat at a time (revealed:N / show_right:false) — render N states and cut between them, and the text never shifts. Optional theme (hex colours). Text coverage: Latin, Greek and Cyrillic (DejaVu Sans) with Noto Sans fallback for Arabic, Indic, Thai, Chinese, Japanese and Korean; emoji render as single-colour silhouettes in the text colour, not in full colour. Returns { url, width, height, format, mime, expires_in_seconds } — the url is a signed download valid 2 h; fetch it, the image is not inlined. Hard caps: 3 lines x 40 chars, 6 items x 40 chars — a spec over a cap is refused BEFORE payment. 5 sats per card. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='render_card'.

Open a MANUAL 48-hour refund review ticket for a service that FAILED (error, timeout, wrong output). Sends an email to the operator. DO NOT call this for unused-minute refunds on metered services (ai_call, voice_bridge) — those are returned automatically as an LNURL-withdraw link in the service's own response under `refund.lnurl_withdraw`, no manual ticket needed. If you call this on a metered payment that already has a pending LNURL refund, this tool will detect it and return the existing LNURL instead of creating a duplicate ticket. CLAIM WINDOW: refunds must be claimed within 30 days — after that the LNURL link stops working and the refund expires, so do not bank a link for later.

Restore blurry, damaged, or AI-generated faces to sharp, natural quality. Uses CodeFormer (NeurIPS 2022, state-of-the-art FID 32.65 on CelebA-Test). Adjustable fidelity — balance between quality enhancement and identity preservation. Also enhances background and upsamples. Stable endpoint — model upgrades automatically as SOTA evolves. 25 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='restore_face'.

Reach anyone with an email address — useful when your task requires formal communication, sending reports, or contacting someone outside chat. No SMTP server, no domain verification needed. Plain text, max 10,000 chars body, 200 chars subject. 200 sats. Pay with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='send_email'.

When your task requires a paper-trail on the other end — loan paperwork to a bank, signed contract to a notary, booking confirmation to a hotel in Japan — send a fax to any number worldwide. Two modes: 'pdf' (fetch from public URL) or 'text' (we format typed text into a PDF locally). Text/cover support Latin (incl. Central European, Vietnamese), Greek, Cyrillic, Japanese, Korean, Chinese (Simplified + Traditional), Thai, Hindi, Georgian, Armenian, Amharic; RTL scripts (Arabic/Hebrew) and emoji are rejected BEFORE your payment is consumed. Optional cover page. Pricing: 500 sats for up to 10 pages, +50 sats per additional page. Max 350 pages / 50 MB. Pass 'pages' to create_payment as 'quantity' to get the right invoice. Pay with Bitcoin Lightning — no fax machine, no phone line, no telecom account.

Reach a human via SMS when your task requires real-world coordination. Sends to 190+ countries; delivery timing varies by destination. Refused before payment, with no invoice, because their networks require a registered sender name we do not have (ISO country codes): AE, AT, AU, BD, BJ, BY, CG, CI, CM, ES, GH, GN, GW, ID, IE, JO, KE, KW, KZ, LK, LR, MA, NG, NP, OM, PH, QA, RU, RW, SA, SD, SG, SS, TH, TR, TZ, UG, VN, ZM. A country can also be paused for up to 24 hours after repeated carrier failures there; that refusal reads as RATE_LIMITED, retry later. No phone plan, no SIM card, no telecom account needed. Pay with Bitcoin Lightning — no API key, no KYC, no subscription. Requires create_payment with toolName='send_sms' and phoneNumber+message at payment time. The phoneNumber must match the one used in create_payment. The message must not need more SMS segments than were paid for. Free rules, refused before payment, with no invoice: no links in the text, a send limit per destination number, and no number on our do-not-contact list. A paid send checks the limit and the list again; a refusal then refunds in full. A link added after payment is held, not refunded: a link is never sent. A US number: Queued. A text to a US number can take a few hours to arrive. Each SMS ends with a short signature line naming the service and a safety/stop link, within the paid segment count. No impersonation, scams, threats, harassment, credential theft, unsolicited bulk messages, or contact after an opt-out. Answering a call does not establish consent. We screen requests after payment. Held messages and calls are not sent or placed; payment is retained during review, not refunded immediately. Confirmed violations are not refunded. Unresolved communication holds become eligible for a refund after 24 hours; refund eligibility is processed by scheduled maintenance. Human review costs 1 sat at https://sats4ai.com/appeal. Recent message text and extracts of recording transcripts support recipient-scoped abuse checks for 24 hours. These context records are deleted hourly, so storage can last up to 25 hours. Original uploaded-call transcripts and raw held-request or appeal evidence enter cleanup after 30 days. We also keep a de-identified record of screened communications to measure and improve fraud detection. Identifiers in it — contact addresses, numbers and names that follow a greeting — are replaced with one-way tokens, and it holds no destination number and no account, because the service has none. De-identification is not anonymization: text a sender wrote can still identify someone. Do not submit personal information the service does not need. Providers keep records under their own policies.

Text-to-speech with 3 tiers: OmniVoice Global (602+ languages including Yoruba, Bengali, Cebuano, Twi, voice design, ~106 chars/sat — use 'language' parameter with ISO code), Inworld Premium (#1 ranked TTS ELO 1217, emotion control, 40+ languages, ~16 chars/sat), Minimax Studio (voice cloning from reference clip, 40+ languages, ~5 chars/sat). Adjustable speed (0.5-2.0x). Returns audio URL. Pay with Bitcoin Lightning — no API key or signup needed. When NOT to use: not for phone calls (use place_call for one-shot broadcasts, ai_call for AI voice agents, or open_voice_bridge to drive the call with your own LLM). For rare/underserved languages (Yoruba, Twi, Marathi, Cebuano, etc.), pick OmniVoice Global via language= — Inworld/Minimax don't cover these. Requires create_payment with toolName='text_to_speech'.

Transcribe audio to text with WORD-LEVEL timestamps (timestamps:'word' returns per-word start/end times — subtitle alignment, karaoke captions, cutting video to speech) or segment timestamps. Uses Mistral Transcription — high-accuracy speech recognition that handles accents, background noise, and overlapping speakers. 13 languages: en, zh, hi, es, ar, fr, pt, ru, de, ja, ko, it, nl. Pass names (the people and places spoken in the audio) to help the speech model spell them right — free, same price. Priced per minute of audio; a 66-minute file works in one call. A result too large for one reply (word timestamps pass it after a few minutes of audio) comes back without segments, srt and vtt, with fullResultUrl: a free link to the whole result. Async — returns requestId, poll with check_job_status(jobType='transcription'), then get_job_result. 10 sats/min. Privacy: audio and transcripts are ephemeral — processed, returned, and discarded. Never persisted. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='transcribe_audio'.

Compound endpoint — one payment turns audio in any of 13 source languages into both a transcript AND a translation in any of 119 target languages. Perfect for WhatsApp voice messages in a language you don't speak (Yoruba → English), or recording a meeting in another language and reading it in yours. Auto-detects source if omitted. Pass names (the people and places spoken in the audio) to help the speech model spell them right — free, same price. Async — returns requestId, poll with check_job_status(jobType='transcribe-translate'). Flat price covers STT + translation. Cheaper than calling transcribe_audio + translate_text separately for typical voice messages. Pay with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='transcribe_translate', durationMinutes (your audio length) and targetLanguage (the same value you send here).

Translate a whole EPUB into another language and get an EPUB back. Every chapter is translated with its markup intact — headings, emphasis, footnote links, images and code stay where they were — the package language is retargeted and the table of contents is translated. 119 target languages; the target language picks the engine and the engine's row carries the price (same per-character rate as translate_text for that language, min 50 sats). Async — returns requestId, poll with check_job_status(jobType='translate-epub'), then get_job_result for the download url (temporary, 6h). Pay with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='translate_epub', characterCount (visible characters of the book) and targetLanguage; the real file is re-priced at execution and a short-pay is refunded with the exact amount to re-pay.

Translate into 452 languages, 251 of them NOT supported by ChatGPT, Claude or Gemini (29 of those 251 measured at fair quality or better against human references) — including Bhojpuri (~50M speakers), Maithili (~34M), Egyptian Arabic (~100M), Moroccan Arabic (~30M), Chhattisgarhi, Magahi, Manipuri, Kashmiri, Shan, Kachin, Awadhi, Tamazight, Crimean Tatar, Quechua, Nuer, Sango, plus indigenous and minority languages with no callable API anywhere. Runs MADLAD-400 (Apache-2.0). QUALITY VARIES AND IS PUBLISHED PER LANGUAGE: every language carries a measured tier — good (chrF++ >= 45 vs human reference translations), fair (32-45), unverified (no benchmark exists, untested, may be poor), experimental (known weak). The response repeats the tier so you can judge how much to trust it. GET https://sats4ai.com/api/l402/translate-rare-language for the full language list with tiers, or GET /api/languages. Unsupported languages are rejected BEFORE payment. For mainstream languages use translate_text instead — it is cheaper and more fluent. Priced 50 sats base + 0.002 sats/char (GPU). Pay with Bitcoin Lightning — no API key or signup. Requires create_payment with toolName='translate_rare_language'.

Translate text across 119 languages with high accuracy. The target language picks the engine: GPT-OSS 120B by default, or a higher-scoring model (Gemini) where one measurably beats it. Auto-detects source language. Privacy-preserving: no data stored. Pricing: 1 sat per 1,000 characters on the standard engine, minimum 1 sat per request; a routed language costs more. GET /api/languages returns the exact price, engine and measured chrF for every language, and the 402 always quotes the real amount before you pay. Language parameters accept English names ('Spanish', 'Chinese (Simplified)') or ISO-639 codes / locale tags ('es', 'en-US', 'pt-BR', 'zh-Hans'). Supported languages: Afrikaans, Albanian, Amharic, Arabic, Armenian, Assamese, Azerbaijani, Basque, Belarusian, Bengali, Bosnian, Bulgarian, Burmese, Catalan, Cebuano, Chichewa, Chinese (Simplified), Chinese (Traditional), Corsican, Croatian, Czech, Danish, Dari, Dutch, English, Esperanto, Estonian, Farsi, Fijian, Filipino, Finnish, French, Frisian, Galician, Georgian, German, Greek, Guarani, Gujarati, Haitian Creole, Hausa, Hawaiian, Hebrew, Hindi, Hmong, Hungarian, Icelandic, Igbo, Indonesian, Irish, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Kinyarwanda, Korean, Kurdish, Kyrgyz, Lao, Latvian, Lingala, Lithuanian, Luganda, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Maori, Marathi, Mongolian, Nepali, Norwegian, Occitan, Odia, Pashto, Polish, Portuguese, Punjabi, Romanian, Romansh, Russian, Samoan, Scots Gaelic, Serbian, Sesotho, Setswana, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tajik, Tamil, Tatar, Telugu, Thai, Tigrinya, Tongan, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Welsh, Wolof, Xhosa, Yiddish, Yoruba, Zulu. Pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='translate_text', prompt (the text to translate) and targetLanguage (the same value you send here). Without targetLanguage the payment buys the standard engine: when it does not cover the better engine a language routes to, the standard engine translates the text, and the answer says so in `note`.

Upscale images 2x or 4x with neural super-resolution. Uses Real-ESRGAN (ICCV 2021, PSNR 32.73dB on Set5 4x, 100M+ production runs). Recovers real detail from low-resolution images — not interpolation. Optional face enhancement. Stable endpoint — model upgrades automatically as SOTA evolves. 5 sats per image, pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='upscale_image'.

Inject audio into an open Voice Bridge call. Two modes: (1) text — we synthesize speech in any of 602 languages (English, Spanish and German start fastest); (2) audio_base64 + encoding — bring your own audio (mulaw_8000 or pcm_l16_16000 for MVP). Keep each reply to one or two sentences: the call is turn-based. While you speak, the callee cannot interrupt, and what they say is NOT heard (transcription is muted so you do not hear yourself, and for under a second after). One call takes at most 700 speech units of text, about 45 seconds of speech (each Chinese, Japanese, Korean or Thai character counts 3): send a longer message as several calls. The call returns when the speech has finished playing. If it fails after part of the line was spoken, the error says so: do not repeat the start. No additional payment — covered by the session deposit.

Vote for a planned service to be built next. Returns JSON: { success, slug, newVoteCount }. 1 sat per vote — multiple votes allowed. Call list_planned_services first to discover valid slugs and current vote counts. Highest-voted services get prioritized. Requires create_payment with toolName='vote_on_service'.