Ornith-1.5 Open-Weight LLMs Highlighted in Posts
Users quote a tweet from @ornith_ on Ornith-1.5 open-weight models with mobile 9B and larger MoE variants.
TLDR
@RuiCarrilho5 stated that 9B models run on mobile and keep up with Qwen 3.6, adding that acceleration cannot be stopped and that by year end phones will run models more powerful than last year's SOTA. @witcheer described a new open-weight model family that ships with Hermes Agent setup, allowing an agent to point at a local Ornith 1.5 server across sizes from 9B to 397B, and noted benchmarks are coming soon. Both posts link to the same tweet from @ornith_.
Combined views
21.6K
3 Sources, first seen ago
Sources
RC 
Roy Carrilho@RuiCarrilho59B models, that can run on mobile now keeping up with Qwen 3.6. the acceleration cannot be stopped. by the end of the year, you'll be able to run models more powerful than last year's SOTA on your *phone* https://twitter.com/ornith_/status/2090074077084127302
- likes: 77
- replies: 4
- bookmarks: 24
- reposts: 4
WI witcheer@witcheera new open-weight model family where the model card ships with Hermes Agent setup out of the box. point your Hermes Agent at a local Ornith 1.5 server, 9B to 397B! benchmark coming soon :) https://twitter.com/ornith_/status/2090074077084127302
- likes: 178
- replies: 15
- bookmarks: 43
- reposts: 14
MR Martin Ronfort@RonfortMartin@RuiCarrilho5 This is exactly that moment. A 9B model running on your phone while keeping pace with Qwen 3.6 desktops. That's not incremental, that's the acceleration you're calling out. We actually covered this here: https://x.com/AGTPinsights/status/2090096806646395279
WI Witcheer@witcheer@Chris_Wozniczek oh that’s cool!
SH Shikhar@xikhar@witcheer I NEED to test this model.
OR