Clef brings vision to document AI, stunning benchmark results go under-hyped
Old-Antelope-4447 · reddit · 2026-10-11
A Reddit post highlights Clef, a new model that extends what Jev does by adding vision support, treating documents as a list of images rather than plain text.
The author claims this enables tasks like checking whether a document is signed or whether an invoice total exceeds a threshold, with calibrated outputs, and says the published benchmark results are astounding. They wonder why the community hasn't shown more excitement about it.
More from Models
- Researcher: Copilot Premium turned one-line data analysis into long-winded prompt begging — lescarr · 2026-10-11
- Berkeley Talk: LLM Reasoning Has Structure; Confidence-Based Stopping Cuts Thinking Tokens by 25% — datawithsuman · 2026-10-11
- Researcher questions whether 'release first, improve later' metrics are worth hillclimbing — suchenzang · 2026-10-11
- Reddit: Google pulls Gemini 3.1 Pro from Antigravity IDE, leaving devs on Flash — Adventurous-Week-399 · 2026-10-11
- Using Grok to scan X for expert takes on OpenAI's 'scary math' and crypto — markjeffrey · 2026-10-11
- User demands: restore limits, 40x usage for $500 sub, and an Opus 5.5-class model — CtrlAltDwayne · 2026-10-11