GPT Sol · 发布历史
OpenAI GPT Sol 系列已收录 4 次发布和 47 个测试版本。
GPT Sol 命名系列,直接前代关系依据官方材料保留。
全部版本
- GPT-6.2 Sol15 项测试
- GPT-6.1 Sol15 项测试
- GPT-6 Sol11 项测试
- GPT-5.6 Sol24 项测试
已收录测试 · 47
- Agents' Last Exam · v1.1; current chart task snapshot and scoring details unspecified; not assumed to be the 2026-10-17 snapshotGPT-6.2 Sol
- Terminal-Bench-Science · 0.1GPT-6.2 Sol、GPT-6.1 Sol、GPT-6 Sol
- Artificial Analysis Intelligence Index · v5.0.1; nine evaluationsGPT-6.2 Sol
- FrontierCode · 2.0GPT-6.2 Sol
- LifeVal · Version and exact task set unspecified in release articleGPT-6.2 Sol
- Harvey LAB-AA · Artificial Analysis v1.1; 120-task held-out set; hallucination-gated and ungated all-pass ratesGPT-6.2 Sol
- GDPval · v2 internal; scoring rubric and exact task snapshot unspecifiedGPT-6.2 Sol
- Vals Science Index · v1GPT-6.2 Sol
- BenchCAD · 2.1 Core; tools and geometric metric not specified in this chartGPT-6.2 Sol
- AutomationBench · 1.0.6; 47 tools; chart-reported score, scoring rule unspecifiedGPT-6.2 Sol、GPT-6.1 Sol、GPT-6 Sol
- AA-3D · Version and leaderboard snapshot time unspecified in release articleGPT-6.2 Sol
- HealthBench Professional · 525 physician-authored conversationsGPT-6.2 Sol、GPT-6.1 Sol、GPT-6 Sol、GPT-5.6 Sol
- MentalHealthBench · Overall; 1,215 synthetic conversations; four independent responses per task; clinician-authored rubricsGPT-6.2 Sol
- GPQA Diamond · 198-question Diamond subsetGPT-6.2 Sol、GPT-5.6 Sol
- User-flagged factuality · Refreshed 1,000 de-identified conversations: 229 retained cases and 771 new cases; separate general factuality and reported-issue graders; not comparable with earlier system-card versionsGPT-6.2 Sol
- DeepSWE · v1.1GPT-6.1 Sol、GPT-6 Sol、GPT-5.6 Sol
- GDP.pdf · OpenAI 2026-09 chart; 10 professional domains; chart score, scoring rule and exact set version unspecifiedGPT-6.1 Sol、GPT-6 Sol
- OSWorld · 2.0 v2026.08.08 offline set; partial reward scoreGPT-6.1 Sol、GPT-6 Sol
- User-flagged factuality, high-difficulty prompts · Anonymized ChatGPT conversations previously flagged by users for factual errors; exact set version and sample size unspecifiedGPT-6.1 Sol、GPT-6 Sol
- HealthBench · 48,000+ rubric items across 5,000 multi-turn conversationsGPT-6.1 Sol、GPT-6 Sol、GPT-5.6 Sol
- HealthBench Hard · System-card evaluation; exact dataset version unspecifiedGPT-6.1 Sol、GPT-6 Sol
- HealthBench Consensus · System-card evaluation; exact dataset version unspecifiedGPT-6.1 Sol、GPT-6 Sol
- Agents’ Last Exam · V1; 55 subfields; exact task snapshot unspecifiedGPT-6.1 Sol、GPT-6 Sol
- MMMU-Pro · Vision format explicitly labeled by the official comparison chart; task release/version unspecifiedGPT-6.1 Sol
- OSWorld · 2.1 offline set; task count and exact task release unspecified in the selected GPT-6.1 Terra article/chartGPT-6.1 Sol
- Factual error rate on difficult prompts · Internal error-inducing prompts in the GPT-6.1 Terra comparison article; exact dataset version, sample size and user-flagged provenance unspecified.GPT-6.1 Sol
- Terminal-Bench · 4.0GPT-6.1 Sol
- Agents’ Last Exam · 55 professional fields; exact task version unspecifiedGPT-5.6 Sol
- GDPval-AA · v2GPT-5.6 Sol
- Big Finance Bench · Version and task protocol unspecifiedGPT-5.6 Sol
- Artificial Analysis Intelligence Index · v4.1GPT-5.6 Sol
- Artificial Analysis Coding Agent Index · v1.1GPT-5.6 Sol
- SWE-Bench Pro · Subset/version unspecified in sourceGPT-5.6 Sol
- Terminal-Bench · 2.1; source metric reported as score percentageGPT-5.6 Sol
- GeneBench Pro · Version unspecified in sourceGPT-5.6 Sol
- LifeSciBench · Version unspecified in sourceGPT-5.6 Sol
- OSWorld · 2.0; subset and scoring details unspecifiedGPT-5.6 Sol
- BrowseComp · 1,266 题;单代理与多代理按协议区分GPT-5.6 Sol
- BenchCAD · Task version and scoring details unspecifiedGPT-5.6 Sol
- BenchCAD · Task version unspecified; Python tool protocolGPT-5.6 Sol
- MMMU-Pro · Standard and Vision formatsGPT-5.6 Sol
- MMMU-Pro · Tool-enabled protocol; exact version unspecified (gpt56-mmmu-pro-with-tools)GPT-5.6 Sol
- GDP.pdf · 100 professional PDFs and promptsGPT-5.6 Sol
- FrontierMath · Tier 1–3 v2GPT-5.6 Sol
- FrontierMath · Tier 4 v2GPT-5.6 Sol
- Toolathlon · Version, subset, and metric unspecified in the 2026-07-09 articleGPT-5.6 Sol
- ARC-AGI-3 · Version/subset unspecified in sourceGPT-5.6 Sol