Skillware framework claims 97.3% cost cut as 3B-8B local models beat frontier LLMs
RossPeili · reddit · 2026-11-06
A new arXiv preprint (arXiv:2611.07652) proposes Skillware, a framework for agent reliability that separates cognitive planning from execution: instead of prompt-engineering models to write code or parse messy text APIs on the fly, models are equipped with pre-vetted, deterministic software bundles. In stress tests across 23 frontier and open-weight models, small local models (3B-8B params) with this architecture systematically beat unassisted frontier models on complex multi-step workflows, with drastically lower token usage — the authors report a 97.3% cost reduction and better reliability. The project is open source (arpahls/skillware on GitHub) with 20k downloads.
More from coding & agent
- Exo: an agent harness that improves itself at runtime, slashing cost and time to SOTA solutions — berkeley_ai · 2026-11-06
- Conductor founder joins Meta, 134 AI coding-agent alternatives flood X to fight for its users — jasonkneen · 2026-11-06
- Anaconda demos CLI + MCP with Kilo Code: managing conda environments via natural language — anacondainc · 2026-11-06
- Search fails new hires: project homes let teams find things they can't yet search for — dSebastien · 2026-11-06
- Running a full experiment suite from a phone via Cursor cloud agents — malliktwts · 2026-11-06
- 10 days with Bend2: migrating core TypeScript logic to functional style with laws and proofs — mattpocockuk · 2026-11-06