Meta's Model Tops Medical Evaluation
Scobleizer · x · 2026-07-14
The core shared update is that Meta's Muse Spark 1.1 performed exceptionally well on the HealthBench Professional medical evaluation. The author claims it achieved a better overall score across 525 real-world clinical tasks and essentially matched GPT-5.6 Sol after length adjustment.
Another key highlight is cost: the model's input/output pricing is listed at $1.25/$4.25 per M tokens, compared to $5/$30 per M tokens for the baseline model, making the output side roughly 7x cheaper. The original poster interprets this as a clear "comeback" for Meta in the AI race.
More from Models
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27