GPT-5.6 Leads in Knowledge Work Benchmark
ArtificialAnlys · x · 2026-07-11
AA-Briefcase is a new agentic knowledge work benchmark designed to test models' real-world capabilities in complex projects, with tasks built by industry experts. Its scoring combines binary Rubric Checks, analysis quality, and pairwise comparisons of presentation effectiveness.
According to the post, GPT-5.6 Sol (max) achieved the highest Presentation Elo on this benchmark, meaning it delivers the best presentation quality across different file types like PowerPoint and Excel. Compared to its predecessor GPT-5.5 (xhigh), its Presentation Elo increased by roughly 500 points, which translates to an estimated 95% win rate in head-to-head visual comparisons.
More from Models
- BullshitBench update: GPT-6-Astra beats all prior OpenAI models but still trails Anthropic — scaling01 · 2026-09-11
- Astra Scores 83% on GauntletBench, First Computer-Use Agent to Beat Human Baseline — ducha_aiki · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- DeepSeek V4 Pro API to continue after Sept 2026, billing unchanged — teortaxesTex · 2026-09-11
- DeepSeek V4.1 Flash Hits 98% of GPT-6 Astra's Score at 1.4% of the Cost in Third-Party Benchmark — ayushtweetshere · 2026-09-11