Dev Radar
Support
LiveUpdated 2026-11-05 16:12 UTC

Compelling little model. Took a look today.

Compelling little model. Took a look today. TwIL-8B isn't meant to be an all-purpose agent brain. It's a math/code…Compelling little model. Took a look today. TwIL-8B isn't meant to be an all-purpose agent brain. It's a math/code…

This is a dev post classified by AGK Intelligence as AI dev tools (an opinion), kept by the Dev Radar because it carries real work, not commentary.

Compelling little model. Took a look today. TwIL-8B isn't meant to be an all-purpose agent brain. It's a math/code specialist, and the card says tool use wasn't trained or evaluated. So I ran the verbose Q4_K_M against NeoHorse 1 9B (my 9B bar) on a 10GB 3080, reasoning off. What transferred surprisingly well for a non-agent model: • Tool calls (T1): 28/30 vs NeoHorse 30/30 • Light briefing (A1): 9.0 vs 10 • Daily organisation (PA1): 54.7 vs 48 (out of 100) • Adding reasoning is slow, but powerful. It lifted PA1 from 54.7 to 79. What didn't: • Context search (C1 @ 16K): 4.75 (vs 9.7). Tha

Posted by GooGZ AI (1.6k followers) 4 h ago · 12 likes · 684 views · view the original post on X. Kept by the Dev Radar as AI dev tools. Tools mentioned: twil-8b.

More dev work like this

Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.6k posts from 5k X accounts over the last 21 days, 9.4k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-05 16:12 UTC. Full method.