NeoHorse-1 debuts: Apache 2.0 4B/9B models trained from agent runs, 262K context
PrajwalTomar_ · x · 2026-09-16
NeoHorse-1 is TokenRhythm's first test of its "learning from agent runs" idea: starting from Qwen3.5, the team trained 4B and 9B versions using data generated from agent runs.
Both models ship with a native 262K context window, can be self-hosted, and use the Apache 2.0 license. The author's takeaway: small models are getting useful fast.
More from Models
- Indie dev open-sources Qwen-2.5-1B-RLCD with 5x faster on-device JSON inference — suchenzang · 2026-09-16
- LLaDA-Image: 6B fully-diffusion DiT trained on 90% image-only data, no caption bottleneck — jiqizhixin · 2026-09-16
- GPT-5.6 Sol flips its conclusions when you just ask 'Are you sure?' — Sockand2 · 2026-09-16
- Student finds Opus and ChatGPT trash his slides, then reverse after seeing the source material — Kazoru4 · 2026-09-16
- Altman says internal post-Astra OpenAI model can solve problems the world's best mathematicians cannot — rohanpaul_ai · 2026-09-16
- Agent randomly references a money quote the user never said while editing unrelated text — Kyrannio · 2026-09-16