Ant open-sources Ling-3.0-flash-Sante: a 124B medical reasoning model with agent tools
aigclink · x · 2026-10-23
Ant Group has open-sourced Ling-3.0-flash-Sante, a medical-augmented model targeting professional medical tasks and complex medical reasoning.
- Architecture: built on Ling-3.0-flash with hybrid attention and sparse MoE, 124B total / 5.1B activated parameters, 256K context
- Agent capabilities: thinking mode, tool calling, and multi-agent collaboration for literature retrieval, comparing clinical studies, evidence synthesis, and long medical document analysis
- Benchmarks (multi-agent): 86.9 on BrowseComp; 83.8 on DiagnosisArena-MCQ, beating Kimi K3 (78.4) and GPT-5.6 Sol (81.9); 53.9 on MedXpertQA-Text, slightly above Kimi K3 but below GPT-5.6 Sol (60.2) and Gemini 3.6 Flash (62.4)
Self-hostable, suited for building medical literature research, clinical guideline retrieval, drug fact-checking, and evidence synthesis applications.
Related event: Ant Group Open-Sources Medical MoE Model Ling-3.0-flash-Sante(7 posts)→
More from coding & agent
- logplane: A Local Control Plane for AI Agents With Recovery, Approval, Budgets — max_paperclips · 2026-10-23
- SlopCodeBench: Top Model Scores Just 16% on Long-Horizon Coding Agent Benchmark — ajratner · 2026-10-23
- 15-year-old ships multi-agent app Lunaar in 4 days, nets 2,000 downloads — Adept_Direction2591 · 2026-10-23
- Charter v0.3: a formally verified language for institutional rules with AI constitutional prompts — QuixiAI · 2026-10-23
- DeepSeek harness slashes tokens, time and cost vs Codex, but tanks on FrontierCode — teortaxesTex · 2026-10-23
- The Real Cost of Bad RAG Retrieval Isn't Tokens—It's the Human Cleanup — Street-Ad4993 · 2026-10-23