Extract
Extract is an entity extraction model from ScaleDown that accepts text, images, and files and returns text. Callers describe the fields they want in plain English rather than choosing from a fixed set of entity types. It has a 1 million-token context window.
Extract specs and pricing
- Provider: ScaleDown
- Released: October 13, 2026
- Context window: 1M tokens
- Max output: 900K tokens
- Input: text, images, files
- Output: text
- Price: $0.04 input / Free output per 1M tokens
What Extract is good at
- Extracting custom fields from text
- Finding named entities in files
- Extracting information from images
To-dos to try with Extract
- Read this text and list every person and their job title.
- Extract the company names and dates mentioned in this file.
- Read this image and list the product names and prices shown.
FAQ
- Who makes Extract? Extract is made by ScaleDown.
- How do I specify what Extract should find? Describe the fields you want in plain English through `provider.options.scaledown.entities`.
- How much does Extract cost? Extract costs $0.04 per million input tokens and Free per million output tokens through OpenRouter. In todo.is a typical agent step costs about 0.11 credits ($1 = 100 credits).
- What is the context window of Extract? Extract can read up to 1M tokens at once and write up to 900K tokens in one reply.
- Can I use Extract without coding? Yes. In todo.is you pick Extract as your agent's model and write to-dos in plain words, on the web, the Android app, WhatsApp or Telegram. No API key needed: credits are included, or you can bring your own key.
JavaScript is required to use the todo.is app.