Dense Slides: Building Cursor Composer – Lee Robinson, Cursor
Source Video
Building Cursor Composer – Lee Robinson, Cursor
Method
This deck is slide-only. The existing captured video frame set supplies candidate frames, then local OpenCV rejects sponsor/title/speaker-only frames, crops visible slide surfaces, deduplicates, and saves the cropped slide images.
Cropped Visible Slides
- Recreated text/layout view: open HTML recreation
- AI slide classifier:
content_slideconfidence0.97 - Text source: advanced OCR
rapidocr-live/bright-screen/contrast. - OCR decision: ready — Dense benchmark chart with small axis labels and multiple text regions; OCR will read this more reliably than direct transcription.
Slide text:
best-in-class speed Composer combines coding intelligence with
Best Open
Fast Frontier
Frontier7/2025
Composer
Best Frontier
0 20 40 60 0 100 200
Composer Intelligence (CursorBench score) Speed (Tokens per Second)
AIE /CODE BUILDING A FAST FRONTIER MODEL WITH RL
Google DeepMind PRESENTED8Y LEE ROBINSON /VP, Developer Education CURSOR
- Recreated text/layout view: open HTML recreation
- AI slide classifier:
content_slideconfidence0.99 - Text source: agent_vision.
Slide text:
Why Build Composer? Inspired by Cursor Tab Our aim was a fast, interactive model that is delightful to use. Experiments with Cheetah Prototype fast model in Cursor, quickly became popular.
Classification audit: raw/sources/slide-ai-classification/dense/fL1iJHtl51Q/audit.json