Early benchmark results put Kimi K3, GPT-5.6 Sol and Claude Fable 5 at 42/42
bdsqlsz · x · 2026-07-21
Early benchmark results compare several models on the same six-part task set.
- Kimi K3†, GPT-5.6 Sol (xhigh)†, and Claude Fable 5 all score 42/42.
- GPT-5.6 Sol (default) drops to 28/42, Meta Muse Spark 1.1 to 26/42, and xAI Grok 4.5 to 13/42.
- Cost differs sharply: roughly $31 for Kimi K3†, $51 for Claude Fable 5, about $21 for GPT-5.6 Sol (xhigh)†, and only $4 for GPT-5.6 Sol (default).
The post says more results are coming soon, implying this is an early snapshot rather than a final leaderboard.
Related event: Multiple Frontier AI Models Achieve Perfect Scores in IMO 2026 Testing(6 posts)→
More from Models
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11