Intern Large Models (Kun Cai, Haodong Chen et al.) · Sci-Agent MoE
Sage S2.1 Flash Technical Report
The report describes Sage Flash, a 320-billion-parameter general-purpose mixture-of-experts model that reasons over scientific evidence and acts in interactive environments. It uses dedicated encoders for molecular graphs, biological sequences and spectra, alongside text and images, and was trained with verifiable environments for tool use, search and coding. Across 33 benchmarks, it ranks first on OfficeQA Pro (72.2), MLE-bench Lite (90.3 HumanRank), Harvey LAB and SpreadsheetBench 2, and second on OSWorld-Verified (85.2) and WebArena-Verified (67.6). The page text refers to the model as Sage S1.5 Flash in places, and does not state licence or availability.
Open the source ↗Similar Language Models releases
- TeleChat to Xing4: Advancing Frontier Agentic Models on Domestic Chips and Frameworks
Xingchen AGI Lab (XingChen-AGI) · Xing 4 LLM · Week 48, 2026
- K2-Horizon
Institute of Foundation Models (K2 Team) · Open Model Fleet · Week 48, 2026
- Qwen4-Max
Alibaba (Qwen team) · Sparse MoE LLM · Week 48, 2026
- LongCat-2.5
Meituan (LongCat team) · Agentic MoE LLM · Week 47, 2026
- Inkling 2 Small
Thinking Machines Lab · MoE agentic LLM · Week 46, 2026
- Ling-3.2-flash (Ling model series)
Ant Group · Multimodal MoE LLM · Week 46, 2026
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.