Ant Ling Releases Medical Model Ling-3.0-flash-Sante

The inclusionAI team under Ant Group released the medical MoE model Ling-3.0-flash-Sante (124 billion parameters, 5.1 billion active, ~262k token context) on September 4, 2026, with free API access and vendor-reported benchmarks, but without published weights.

Compiled by Sergey KostenchukPublished 2026-09-08Updated 2026-09-18

2026-09-08 Research Google

🤖 Ant Ling Releases Medical Model

The inclusionAI team (Ant Group) released Ling-3.0-flash-Sante on September 4, 2026 — a medical variant of Ling-3.0-flash: MoE 124 billion parameters (5.1 billion active), ~262k token context. The announcement claims DiagnosisArena-MCQ 83.8 versus 81.9 for GPT-5.6 Sol, but the figures are vendor-reported and weights are not published.

🌍 Free API access lowers the barrier to entry for clinical applications: medical reasoning and long document analysis without self-hosting a 124B model. The absence of weights and a technical report limits production use.

👤 The model is available for free: OpenRouter (ling-3.0-flash-sante:free) and Vercel AI Gateway until October 4, 2026. The ~262K context is convenient for long medical documents; figures from the announcement should be re-verified on your own prompts.

Source 1: https://x.com/AntLingAGI/status/2095953758148853892 Source 2: https://openrouter.ai/inclusionai/ling-3.0-flash-sante:free

Sources