Datalab has released lift, a 9B open-weights vision model for structured extraction. You pass it a JSON schema, and it returns a JSON object that matches. The model reads PDFs and images directly, ...
The editor knows to read it. An AI agent reading the raw JSON doesn't. So when a founder hands that JSON to an agent and asks it to rebuild the app in code, the agent reads 50 user fields and rebuilds ...
There was an error while loading. Please reload this page.