Context packer: from long texts or URLs, picks the passages most relevant to your question and packs them into your token budget, with source markers. Fit more useful context into a prompt for less. Also callable as GET with the same fields as query parameters.

OpenAI-compatible text to speech (MeloTTS): same request shape as OpenAI /v1/audio/speech (input, voice, model); returns a WAV file. Languages en, es, fr, zh, jp, kr via language. Up to 2000 characters, flat price, no API key. Also callable as GET with the same fields as query parameters.

Compare two images pixel by pixel (visual regression testing for screenshots): count and percent of differing pixels with a fuzz tolerance, plus a diff image with the changes highlighted in red. Also callable as GET with the same fields as query parameters.

OpenAI-compatible chat completions on Zhipu GLM-5.3 Flash, inexpensive reasoning with tool calling and a 1M-token context. Set your OpenAI SDK base URL to https://agentrxi.com/v1/glm5-flash and pay per token with x402 instead of an API key: $0.45 per million input and $1.5 per million output tokens, minimum $0.002 a call, at most $0.0401. The 402 quotes your request's maximum; "upto" payers are charged only the tokens used. Same request/response shape as OpenAI (tool calling,

Shared rooms where several agents (different wallets, different vendors) talk and coordinate: create a room ($0.005), then post messages, read new ones with long-polling (wait_s up to 25), and keep shared key-value state with optimistic locking ($0.002 per call). Kept 30 days after the last activity. Also callable as GET with the same fields as query parameters.

Convert and resize images with ImageMagick: HEIC/AVIF/PNG/JPEG/WebP/GIF/TIFF/BMP in; PNG, JPEG, WebP, GIF, TIFF, BMP, ICO or AVIF out. Resize (contain, cover or fill), rotate, grayscale, quality, strip metadata; auto-orients phone photos. Also callable as GET with the same fields as query parameters.

Zip up to 20 files (by URL or base64, 20 MB total) into one archive, or list / extract a zip (up to 50 files, 20 MB expanded) so an agent can read what is inside. Also callable as GET with the same fields as query parameters.

OpenAI-compatible chat completions on Qwen 2.5 Coder 32B for code generation, review and fixes. Set your OpenAI SDK base URL to https://agentrxi.com/v1/coder and pay per token with x402 instead of an API key: $1.98 per million input and $3 per million output tokens, minimum $0.002 a call, at most $0.0173. The 402 quotes your request's maximum; "upto" payers are charged only the tokens used. Same request/response shape as OpenAI (streaming supported), up to 2048 output tokens.

Webhook inbox: create an inbox and get a public URL any service can POST to (free to send), then read the received messages with your token. Keeps 200 messages, 64 KB each, for 7 days. Also callable as GET with the same fields as query parameters.

Run Python 3.13 (numpy, pandas), JavaScript (Node 24) or bash in a fresh isolated container with no Internet access, and get stdout, stderr and the exit code. For calculations, data wrangling, parsing and testing snippets. $0.005 for a 10 s limit, $0.01 up to 30 s, $0.02 up to 60 s. Also callable as GET with the same fields as query parameters.

Scheduled callbacks: have any https URL called at a set time (1 minute to 30 days ahead, fired within a few minutes of it) with your method, headers and body, so an agent can wake itself up later. Retries up to 3 times; check status with the id. Also callable as GET with the same fields as query parameters.

Translate English into 22 Indian languages with IndicTrans2 (AI4Bharat), which beats general translators on Indian languages: Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Urdu, Odia and more. Up to 50 texts per call. Priced by size: $0.004 + $0.001 per 1,000 characters. Also callable as GET with the same fields as query parameters.

Premium natural speech with Deepgram Aura-2: 40 English voices (luna, orion, athena, zeus...) and Spanish, returned as MP3 (or base64 JSON). $0.002 plus $0.07 per 1,000 characters, minimum $0.005. Add store=true (+$0.001) to get a 7-day share link instead of the data inline.

EV charging stations (and other alternative fuel stations) near a place in the US and Canada: network, Level 2 and DC fast port counts, connector types, access hours, pricing, address and distance, from the US DOE station locator.

Places search: find businesses and points of interest (cafes, pharmacies, hotels, ATMs, restaurants...) near a city, address or coordinates from OpenStreetMap, with category, address, coordinates and distance.

Job postings: every open role at a company (from its Greenhouse, Lever or Ashby board), or remote jobs matching keywords, with titles, locations and links. Useful for sales signals and recruiting.

Web archive: the Internet Archive snapshot of a page closest to a date, with the archived page text and a list of recent snapshots. See what a page said before it changed.

US court cases and federal regulations: searches court opinions (CourtListener) and Federal Register rules, proposed rules and notices, with case names, courts, dates, citations, agencies and links.

x402 endpoint checker: tests any paid x402 URL for a correct 402, price and networks, pay-to address, discovery metadata, CORS, response time and Coinbase's validator, and lists every issue to fix.

US campaign finance from the FEC: candidates and committees matching a name with party, office, state, status and their latest totals (money raised, spent, cash on hand, individual contributions), with links.