← leaderboard
GPT-5.6 Sol
NOT COOKEDOpenAI · openai/gpt-5.6-sol · verdict as of 2026-08-10
Daily test history (90 days, baseline band = trailing mean ± 2σ)
Creative Writing
no data
Summarization
NOT COOKED
Math & Reasoning
NOT COOKED
Code Generation
NOT COOKED
Instruction Following
NOT COOKED
Structured Extraction
NOT COOKED
Report Analysis
NOT COOKED
Customer Service
NOT COOKED
Web Design
NOT COOKED
Game Design
no data
Serving providers (last 30 days)
striped = multiple providers served this model that day (hover for detail)
Recent samples (latest run, one per test case)
No samples from the latest run.