AI models & inference

Deterministic JSON validation and conservative repair for AI-agent payloads. Repairs unquoted object keys, single-quoted strings, trailing commas, Markdown code fences, UTF-8 BOMs and JSON-compatible Python literals, including nested or repeated defects. No LLM or invented data; every applied transformation is reported.

OpenAI-compatible chat completions gateway — standard /v1/chat/completions path and request shape, model chosen in the body: any gpt-* chat model (not Claude/embeddings), each at its own price. Priced per model, $0.005–$0.65 per call in USDC via x402, no API key: POST unpaid and the 402 quotes your model's exact price. Supports response_format json_object. GET /v1/models (free) lists every model with its endpoint and price.

OpenAI-compatible chat completions paid per call in USDC: point any OpenAI SDK at this base URL and send a normal chat request, with the model chosen from a curated list and streaming supported. Use it when an agent needs a model answer without an API key or an account, one call at a time, with a receipt for each.

Anthropic Messages API over x402 - point the Anthropic SDK (or Claude Code / the Agent SDK) at base_url https://agent402.tools/v1/nano and pay $0.003 per call in USDC, no API key, no signup. Same models, caps and price as this tier's /chat/completions route; any model here is served through the Messages wire (Claude natively, others translated). Depth: `effort` (low..max) on Claude 4.7+, thinking.budget_tokens on older Claude. Up to 12,000 input chars and 768 output tokens; streaming supported.

Generate a short video clip from a text prompt at a flat price per clip: a four-second 720p clip returned inline as base64 MP4, with the duration and resolution locked so the price never moves. Use it when an agent needs motion rather than a still, and needs to know the cost before it calls.

Batch text-to-speech: synthesize up to 20 separate text items in a single call, standard voices only, up to 500 total characters, returns JSON with base64 WAV audio plus duration and sample rate per item. Use it to: voice a list of short UI strings at once, generate audio for multiple notification templates, batch-produce voice lines for a menu or form, synthesize several short replies in one request. USDC on Base via x402.

Together AI pricing table: per-token cost of every AI model Together AI serves, in one call — input, output, cache and batch USD per 1M tokens, ranked cheapest first, with context window and capability flags joined in. Compare LLM token cost inside Together AI and against other providers hosting the same model. Normalized from public sources, cross-checked, refreshed hourly.

Attested LLM inference: gpt-4o-mini chat completion plus a post-quantum signed, Hedera-anchored attestation binding the prompt hash, response hash, exact model version, and timestamp. Evidence of which model said what, when — for agents whose decisions may be questioned later.

LLM chat completion (OpenAI-compatible, streaming supported)

American Community Survey 5-year demographics for Scotts Bluff County and each of its ten places, plus a 16-year trend series: population, median household and per-capita income, poverty, median home value, gross rent, housing units and owner/renter occupancy. GET /v1/census.

Use when an agent needs chat completion. Returns One bounded chat completion (standard messages request shape) per paid call, with no account to open and no provider key to hold; max 24,000 input characters and 600 output tokens. $0.01.

Translate text between languages — text translation API for short content: translate a sentence, chat message, or up to 500 words to English, Spanish, French, Chinese, or any of 30+ languages. Language translation with auto-detected source language, formatting preserved; returns the translation plus detected source. Machine translation via Claude Haiku for agents localizing content cheaply. For longer documents use /translate/long.

Call OpenAI's gpt-4.1-mini via a single pay-per-call x402 endpoint — no OpenAI account or API key needed, pay $0.005 per request in USDC. Standard OpenAI chat.completions request/response shape, capped input and output.

Sentiment analysis API — analyze sentiment of text and get a text sentiment score in one call: classifies positive / negative / neutral / mixed polarity with a -1 to +1 sentiment score, plus emotion detection in text (joy, anger, sadness, fear, surprise, disgust, trust, anticipation). Aspect-based sentiment and opinion mining for customer feedback analysis — analyze reviews, support tickets, social posts, chat messages. Via Claude Haiku.

Classify text into your own categories — zero-shot: supply 2–20 labels and Claude Haiku returns the best-matching category with confidence and per-label scores. Route, tag, triage, and detect intent or topic with your own taxonomy in one call.

OpenAI-compatible chat completions gateway (alias of /v1/chat/completions for api/v1-style base URLs) — standard request shape, model chosen in the body: any gpt-* chat model (not Claude/embeddings), each at its own price. Priced per model, $0.005–$0.65 per call in USDC via x402, no API key: POST unpaid and the 402 quotes your model's exact price. Supports response_format json_object. Free model catalog: GET /v1/models.

OpenAI-compatible chat completions with the model chosen server-side: omit model and the gateway routes the prompt to the top-ranked model for its task (code, reasoning, long-context, general) from a fixed eval-derived ranking, failing over automatically on provider errors. Flat price per call, 16k chars in, 1024 tokens out, streaming supported. Use it as a drop-in OpenAI base_url when you want good answers without picking a model.

The OpenAI Responses wire priced from the request itself: each 402 quotes this exact body, from $0.001, and upto buyers settle actual usage under it. Use it when an agent built on the Responses API or the OpenAI Agents SDK wants per-call payment sized to what it sends.

LLM inference proxy (Pro tier) - GPT-4o or GPT-4.1. Supports vision (up to 2 image URLs) and structured output (response_format: json_object or json_schema). No API key needed; pay per call via x402. Input capped at 16k chars, output at 2048 tokens.

Higher-fidelity text-to-image on the OpenAI images wire: a prompt in, one 1024x1024 image out as inline base64, served by a pro-grade diffusion model in about ten seconds. Use it when an agent needs a finished picture for a page, a post or a product and quality matters more than the cheapest draft.

Open-weight text embeddings at a fixed model and precision: deterministic vectors for RAG and memory pipelines, identical across suppliers. POST /v1/text with {inputs} (a string or array of strings) and get the embedding vectors as one JSON object. Pay per request in USDC; no account, no API key.

Text and embedding similarity scoring: cosine match and dedupe decisions at fixed model precision. POST /v1/semantic with {text_a, text_b} and get a similarity score as one JSON object. Deterministic. Pay per request in USDC; no account, no API key.

Image generation: a prompt in, one image back in the same response. 7 models including Google Nano Banana and Nano Banana Pro, OpenAI GPT Image 2, Seedream 5.0 Pro and FLUX.2 Pro, square to 21:9, up to 4K, from $0.02. GET this URL for every model and price. No account or API key: the payment is the only credential.

OpenAI Responses API over x402 - point the OpenAI SDK's responses.create() (or the OpenAI Agents SDK) at base_url https://agent402.tools/v1/auto and pay $0.01 per call in USDC, no API key, no signup. Same models, caps and price as this tier's /chat/completions route; any model here is served through the Responses wire.

Generate an image from a text prompt at a flat price per picture, no token math and no subscription: the OpenAI images request shape in, base64 PNG or JPEG out, with an automatic fallback model so a provider outage does not become your error. Use it when an agent needs a picture cheaply and predictably.

Convert text to speech with Kokoro-82M, a fraction of the price of /api/tts. Returns base64-encoded mp3 or pcm. The same request shape and the same ten voice names as /api/tts, mapped to Kokoro's own voices; the voice is synthetic-sounding where the ElevenLabs tiers are not, which is the whole trade. Use this for high-volume narration, notifications and agent speech where the cost per call matters more than the timbre; use /api/tts or /api/tts-hd when it does not.

Delx Commerce budget image generation API for $0.005 USDC (FLUX Schnell entry tier). Call when you need a cheap image, quick concept, thumbnail, or low-cost draft under the $0.01 draft SKU—paid commercial Replicate only (never mediagen). Public WebP URL + SHA-256 + explicit markup via x402 on Base or Solana.

YouTube transcript API for AI agents: send a video URL or 11-character id and get its published subtitles as timed JSON segments, plain text, SRT or WebVTT, with title, channel, duration, views and publish date. Choose one of ten subtitle languages or any. No speech-to-text. A video without subtitles in that language is not charged.

Find pages semantically similar to a given URL.

Claude pricing table: what every Claude model costs per token right now — Claude 5, Claude Opus, Claude Sonnet, Claude Haiku — from Anthropic and every provider that resells or hosts them (Bedrock, Azure, Vertex, OpenRouter, Fireworks). Input, output, cache and batch USD per 1M tokens, sorted cheapest first, so you can compare Claude inference cost across hosts and pick the cheapest place to run one. Normalized, cross-checked, refreshed hourly.