Expand launches Expand Fetch web scraping API at $1 per 1,000 fetches
Article timeline
The articles in this story. Times are ET.
dex@dexhorthyXDex Horthy praises team's work in short X postAIDex Horthy (@dexhorthy) posts a brief congratulation to his team, "extra dope nice work team," with no product or details given. The post quotes Tim Suchanek's announcement of Expand Fetch, which says its Iris and X-Ray models score 74% on a 1,094-question benchmark versus 55% for the next best API. Expand Fetch costs $1 per 1,000 fetches.
Rohan Paul@rohanpaul_aiXExpand Fetch launches a web-reading API for agents, claiming 74% accuracy on a hard benchmarkAIExpand has released Expand Fetch, a web scraping and crawling API that uses its Iris vision model and X-Ray JSON model to keep tables, icons and hidden page data. On its MDBench Hard test of 1,094 questions across 326 pages, agents using Expand answered 74% correctly, against 55% for the next best API. The product costs $1 per 1,000 fetches and stores each fetch with a Wayback Machine snapshot so answers can be traced to their source.
Santiago@svpinoXExpand's MCP server lets agents fetch web pages as MarkdownAIExpand's MCP server connects agents to the web: add it, sign in when prompted, and ask the agent to use it for fetching pages. Each fetch returns Markdown that keeps the original format, including tables, icons and layout. The server is part of Expand Fetch, which the company launched with an API priced at $1 per 1,000 fetches.
elvis@omarsar0XExpand Fetch reads rendered pages to cut web agent hallucinationsAIExpand Fetch, from Tim Suchanek's Expand, reads rendered web pages and returns underlying data as clean JSON, Elvis Saravia writes. Flattening pages to text can drop details such as a small ✗ in a pricing table, leading agents to answer incorrectly. Expand says its agents answered 74% of 1,094 questions across 326 pages correctly, versus 55% for the next best API, at $1 per 1,000 fetches.