SMB Bench adds six models

Website

• The SMB Bench report leaderboard now scores 15 models: Claude Fable 5.2, Muse Spark 1.3, Muse Ember, Claude Haiku 5.5, GPT-6 Luna and Gemini 4 Carbon join the first nine. The analyses below the leaderboard stay as of the 6 November report.
• Inference costs in the report are shown at each provider's regular rate.
• Run counts and the harness build the six ran on are documented in the dataset card.