AI Benchmark Digest — 2026-07-13

View on AI Benchmark Hub

Daily

New Benchmarks (4)

• LLM Stats (Artificial Analysis) (Score (%)): GPT-5.6 Sol leads with 59.0 across 5 models.
• LLM Stats (AutomationBench) (Score (%)): GPT-5.6 Sol leads with 18.1 across 5 models.
• LLM Stats (DeepSWE) (Score (%)): GPT-5.6 Sol leads with 72.7 across 7 models.
• LLM Stats (GDP.pdf) (Score (%)): Claude Sonnet 5 leads with 81.6 across 5 models.

New Scores From Top-10 Models (10)

• GPT-5.6 Sol on Agents' Last Exam: 30.6 Pass Rate (%) (#1/20)
• GPT-5.6 Sol on DeepSWE: 72.7 Pass@1 (%) (#1/17)
• GPT-5.6 Sol on LLM Stats (GDPval-AA): 1747.8 ELO (#2/38)
• GPT-5.6 Sol on LLM Stats (Graphwalks BFS >128k): 77.1 Score (%) (#2/11)
• GPT-5.6 Sol on LLM Stats (HealthBench Hard): 33.1 Score (%) (#2/9)
• GPT-5.6 Sol on LLM Stats (HealthBench Professional): 60.5 Score (%) (#2/8)
• GPT-5.6 Sol on LLM Stats (HealthBench): 57.0 Score (%) (#3/8)
• GPT-5.6 Sol on LLM Stats (MRCR v2 (8-needle)): 73.8 Score (%) (#3/13)
• GPT-5.6 Sol on LLM Stats (Toolathlon): 58.0 Score (%) (#2/27)
• GPT-5.6 Sol on ZeroEval GPQA Diamond: 94.6 GPQA Diamond Score (#1/230)

New #1 Leaders (4)

• Design Arena (Video Editing) (Elo): Gemini 2.0 Flash (1402.0) beat happy-horse-1.0 (1318.0) by 84.0.
• LLM Stats (Terminal-Bench 2.1) (Score (%)): GPT-5.6 Sol (88.8) beat Claude Fable 5 (84.3) by 4.5.
• WeirdML (Average Score): Claude Fable 5 (Max) (91.94) beat GPT-5.6 Sol (High) (88.76) by 3.18.
• LLM Stats (BrowseComp) (Score (%)): GPT-5.6 Sol (90.4) beat GPT-5.5 Pro (90.1) by 0.3.