LaunchAI ModelsOctober 10, 2026

Microsoft releases Decision-1, a Qwen3.5-9B decision-scoring model

Read original source →vercel.com

Microsoft-Decision-1 returns a calibrated probability per fixed answer option instead of generated text, hitting 83.5% average accuracy across 36 benchmarks at 85 ms p50 latency. It is post-trained from Alibaba's Qwen3.5-9B and priced at $0.042 per million input tokens with free output.

How this story unfolded

1 day · 2 reports · 2 community posts · 4 of 5 shown

  1. Oct 9
  2. Oct 10

More stories today

Open the live feed