NousResearch's Hermes adapter lifts Qwen3.8-27B agent scores: SWE-Bench Pro 29.4%→43.1% at 4× fewer tokens
yb2698 · x · 2026-11-19
NousResearch announced its third lightweight Hermes adapter (after Pi and OpenCode), this time for Alibaba's Qwen3.8-27B, making it a stronger and more efficient agent behind Hermes Agent. Key numbers:
- SWE-Bench Pro (hard 51) pass@1: 29.4% → 43.1%; with xhigh thinking: 45.1% → 51.0%
- Terminal Bench 2.1 pass@1: 44.7% → 58.8% (58.8% → 57.6% with xhigh thinking)
- With thinking off, it matches the base model at xhigh while using 3.5–4.4× fewer tokens
The adapted checkpoint is live on Hugging Face (harness-hub/Qwen3.8-27B-hermes-adapted), demonstrating that an adapter layer alone can meaningfully boost agent performance and efficiency without retraining.
Related event: Nous Releases RL-Trained Hermes Adapter for Qwen3.8-27B(2 posts)→
More from coding & agent
- Every on Claude Fable 5.2: Human Thinking at the Start Matters More as AI Builds More — every · 2026-11-19
- Semantic caching LLM answers by embedding similarity is unsafe — 138-pair test shows why — Narrow_Side_7778 · 2026-11-19
- Can ComfyUI refine existing mocap animations? A motion-to-motion workflow question — _Tito5_ · 2026-11-19
- Mycelium: an open-source framework that turns analytical project knowledge into a living dataset — arjunrajlab · 2026-11-19
- harness-hub open-sources RL-trained LoRA adapters tuned per agent harness, recipe coming Friday — yb2698 · 2026-11-19
- RL-Trained Hermes LoRA on Qwen3.8-27B Adds +13.7 Pts on SWE-Bench Pro While Using Fewer Tokens — yb2698 · 2026-11-19