Shanghai AI Lab open-sources Sage-A1.5: 35B MoE that runs locally, 61.2% on SWE-Bench Pro
vllm_project · x · 2026-10-22
Shanghai AI Lab released Sage-A1.5, an open agentic model with Day-0 vLLM support: vllm serve InternScience/Sage-A1.5.
Key facts:
- 35B MoE with only 3B active parameters
- Agentic orchestration & verification: spawns parallel subagents and self-corrects on the fly
- Trained via an RSI data flywheel built on real scientific discovery workflows and tens of thousands of synthesized environments
- Beats frontier giants on FS-Olympiad, FS-Research, CMT & PinchBench; scores 61.2% on SWE-Bench Pro
- Runs locally via vLLM, SGLang, llama.cpp and LM Studio (Q4KM 20GB)
Weights are on Hugging Face and ModelScope.
Related event: Shanghai AI Lab Open-Sources Sage-A1.5, a 35B MoE Agent Model(6 posts)→
More from coding & agent
- Anthropic unveils Claude Duet, an agentic Claude that does work before you ask — Tegadesigns · 2026-10-22
- 96.5% deliverability, 42.7% opens: why you can't vibe code a newsletter platform — marclou · 2026-10-22
- New Obsidian plugin turns notes app into an Eagle-style asset manager — jiayuan_jy · 2026-10-22
- One-press Hermes agent: open-source ConcoDispatch turns iPhone Action Button into an AI launcher — Teknium · 2026-10-22
- 6 workflows where decision model Jev cuts LLM bills up to 87% — femke_plantinga · 2026-10-22
- Scrawl: open-source Mac tool turns screen markup and voice into agent feedback, 680 stars — aronchick · 2026-10-22