Artificials
Compare models
All makers

AI model maker

OpenAI

107 models on Artificials, 53 with published test results and 5 with open weights you can download.

RSS feed
Best on the capability index166.4GPT-6 Astra, #2 of 274
Top score on10 testsof the 28 its models are tested on
Newest releaseGPT-6.1 SolSep 29, 2026
Models10753 tested, 5 with open weights

Where OpenAI stands

Its best model’s rank among all models tested on each skill area’s main test, the same tests model pages use. A longer bar means a better rank.

  • OverallGPT-6 Astra on Epoch Capabilities Index
    Top 1%
  • CodingGPT-5.6 Sol (max) on LiveBench coding
    Top 8%
  • MathGPT-5.5 Pre-release (xhigh) on Mock AIME 2024–2025
    #1 of 198
  • KnowledgeGPT-6 Astra (max) on SimpleQA Verified
    #1 of 77
  • AgentsGPT-6 Astra (max) on LiveBench agentic coding
    Top 32%
  • LanguageGPT-6.1 Sol (max) on LiveBench language
    Top 4%
  • Data analysisGPT-6 Astra (max) on LiveBench data analysis
    #1 of 63
  • ReasoningGPT-6 Astra (max) on GPQA Diamond
    #1 of 222
  • Image generationgpt-image-1.5 on T2I-CoReBench
    Top 8%

On the capability index

Its best 12 of 46 models on the Epoch Capabilities Index, each with its rank among all 274 models. The top score is 167.3, held by Claude Opus 5.5 from Anthropic.

Every test

Its best result on each published test, and that result’s rank among every model tested. Scores from different tests are never combined.

TestBest modelScoreRank
Epoch Capabilities IndexGPT-6 Astra166.4#2 of 274
GPQA DiamondGPT-6 Astra (max)95.8%#1 of 222
Mock AIME 2024–2025GPT-5.5 Pre-release (xhigh)100.0%#1 of 198
Chess puzzlesGPT-6 Astra (max)72.0%#1 of 141
MATH Level 5GPT-5 (high)98.1%#1 of 97
FrontierMath Tiers 1–3GPT-6 Astra (max)93.7%#1 of 81
SimpleQA VerifiedGPT-6 Astra (max)75.6%#1 of 77
Arena codingGPT-6 Astra (max)1543#5 of 384
FrontierMath Tier 4GPT-6.1 Sol (max)100.0%#1 of 63
LiveBench data analysisGPT-6 Astra (max)83.0%#1 of 63
LiveBench reasoningGPT-6 Astra (max)92.7%#1 of 63
Arena WebDevGPT-6 Astra (max)1788#2 of 123
Arena instruction followingGPT-5.6 Sol (xhigh)1490#9 of 389
LiveBench languageGPT-6.1 Sol (max)90.1%#2 of 63
Arena hard promptsGPT-5.6 Sol (xhigh)1511#14 of 389
DeepSWEGPT-6 Astra (xhigh)74.1%#1 of 26
Arena mathGPT-5.51500#15 of 372
Arena creative writingGPT-5.6 Sol (xhigh)1468#16 of 387
Arena textGPT-5.6 Sol (xhigh)1484#17 of 389
LiveBench mathematicsGPT-6.1 Sol (max)96.8%#3 of 63
SWE-bench VerifiedGPT-5.5 Pre-release (xhigh)80.6%#2 of 32
LiveBench overallGPT-6 Astra (max)82.2%#4 of 63
T2I-CoReBenchgpt-image-1.578.2%#3 of 40
LiveBench codingGPT-5.6 Sol (max)83.9%#5 of 63
LiveBench instructionsGPT-6 Astra (max)75.6%#8 of 63
SWE-bench Verified, bash onlyGPT-5.2 (high)72.8%#7 of 42
LiveBench agentic codingGPT-6 Astra (max)57.3%#20 of 63
CursorBenchGPT-5.6 Sol (max)41.7%#7 of 14

Every model

Newest first. The price is per 1M tokens, blended three input tokens to one output token: OpenAI’s own listing, or else the middle price across hosts.

ModelReleasedContextPriceCapability index
GPT-6.1 SolSep 29, 20261.1M tokens$4.00166.1
GPT-6.1 Sol ProSep 29, 20261.1M tokens$4.00Not on the index
GPT-6 SolSep 22, 20261.1M tokens$4.00162.7
GPT-6 LunaSep 22, 20261.1M tokens$0.20156.3
GPT-6 Luna ProSep 22, 20261.1M tokens$0.20Not on the index
GPT-6 Sol ProSep 22, 20261.1M tokens$4.00Not on the index
GPT Astra LatestSep 11, 20261.1M tokens$20.00Not on the index
GPT Luna LatestSep 11, 20261.1M tokens$0.20Not on the index
GPT Terra LatestSep 11, 20261.1M tokens$4.50Not on the index
GPT Image 2.5 FlareSep 8, 2026Not reported$11.25Not on the index
GPT Image 2.5 SunburstSep 8, 2026Not reported$11.25Not on the index
GPT-6 AstraSep 4, 20261.1M tokens$20.00166.4
GPT-6 Astra ProSep 4, 20261.1M tokens$20.00Not on the index
Daybreak BlueAug 7, 20261.1M tokens$8.00Not on the index
Daybreak RedAug 7, 2026400K tokens$28.13Not on the index

Models, prices, context windows and release dates from models.dev (MIT). Test results from LiveBench (Apache-2.0), Epoch AI (CC BY 4.0), T2I-CoReBench (CC BY-SA 4.0), Datacurve DeepSWE, compiled by Epoch AI (CC BY 4.0), Cursor CursorBench, compiled by Epoch AI (CC BY 4.0), Arena (CC BY 4.0) and SWE-bench (CC BY-NC 4.0). A model counts here when OpenAI sells it, several hosts list it or a test covers it. Each figure is the source’s own; nothing is combined into a new score.