Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

SuperGPQA leaderboard — the family board.

2 TRACKED VARIANTS · 1 FIRST-CLASS BOARD

One family, 2 tracked prints. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA59 MODELS ON THIS BOARDLEADER QWEN3.7 MAX AT 73.6

The board — SuperGPQATOP 59 OF 59 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen
01Qwen3.7 MaxAlibaba73.62026-08-23
02GPT-5.5OpenAI72.72026-09-28
03Qwen3.6 PlusAlibaba71.62026-08-23
04Qwen3.7 Plus PreviewAlibaba71.42026-08-23
05Seed 2.1 Pro PreviewByteDance70.82026-09-28
06Claude Opus 4.5Anthropic70.62026-06-15
07Qwen3.5 397B A17BAlibaba70.42026-08-23
08Seed2.1ByteDance67.42026-09-28
09Qwen3 Max (Reasoning)Alibaba67.32026-08-23
10Qwen3.5 122B A10BAlibaba67.12026-08-23
11Qwen3.6 27BAlibaba662026-08-23
12Gemma 4 31BGoogle65.72026-06-15
13Qwen3.5 27BAlibaba65.62026-08-23
14Seed 1.8ByteDance64.82026-09-09
15Qwen3.6 35B A3BAlibaba64.72026-08-23
16Qwen3 VL 235B A22B ReasoningAlibaba64.32026-08-23
17Qwen3.5 35B A3BAlibaba63.42026-08-23
18Qwen3 235B A22B Instruct 2507Alibaba62.62026-08-23
19Gemma 4 26B A4BGoogle61.42026-06-06
20Qwen3 Next 80B A3BAlibaba60.82026-08-23
21Qwen3 VL 235B A22B InstructAlibaba60.42026-08-23
22Qwen3 VL 32B ReasoningAlibaba592026-08-23
23Qwen3 Next 80B A3B InstructAlibaba58.82026-08-23
24Qwen3.5 9BAlibaba58.22026-08-23
25Qwen3 30B A3B 2507 ThinkingAlibaba57.32026-06-15
26Kimi K2 InstructMoonshot57.22026-08-23
27Claude Opus 4Anthropic56.52026-06-15
28Qwen3 VL 30B A3B ReasoningAlibaba56.42026-08-23
29Claude Sonnet 4Anthropic55.72026-06-15
30Seed Oss 36B InstructByteDance55.72026-06-15
31Qwen3 VL 32B InstructAlibaba54.62026-08-23
32DeepSeek-V3DeepSeek53.72026-06-15
33Qwen3 VL 30B A3B InstructAlibaba53.12026-08-23
34Qwen3.5 4BAlibaba52.92026-08-23
35Hy3-preview-BaseTencent51.62026-06-05
36Qwen3 VL Thinking (8B)Alibaba51.22026-08-23
37Kimi K2 BaseMoonshot51.12026-08-23
38Qwen3 235B A22BAlibaba50.22026-08-23
39GLM-4.5-BaseZ.ai49.62026-06-05
40Qwen3 32BAlibaba49.32026-06-15
41Claude 3.5 SonnetAnthropic48.22026-06-05
42Qwen3 VL 4B (Reasoning)Alibaba46.82026-08-23
43DeepSeek-V3-BaseDeepSeek46.22026-06-05
44o1-miniOpenAI45.22026-06-05
45Qwen3 VL 8B InstructAlibaba44.52026-08-23
46QwQ 32B PreviewAlibaba43.62026-06-05
47DeepSeek-V3.2-Exp-BaseDeepSeek43.62026-06-13
48GPT-4oOpenAI42.42026-06-05
49DeepSeek-V3.1-BaseDeepSeek42.32026-06-13
50MiMo V2 Flash BaseXiaomi41.12026-06-13
51Step3.5 Flash BaseStepFun412026-06-05
52DeepSeek-R1-Distill-Qwen-14BDeepSeek40.62026-06-05
53MiMo 7B RLXiaomi40.52026-06-05
54Qwen3 VL 4B InstructAlibaba40.32026-08-23
55Qwen3.5 2BAlibaba37.52026-08-23
56DeepSeek-R1-Distill-Qwen-7BDeepSeek28.92026-06-05
57Qwen2.5 7B BaseAlibaba26.32026-05-03
58Qwen3.5 0.8BAlibaba21.32026-08-23
59Qwen3 4B Base DenseAlibaba20.92026-05-03
◆ IV — INDEPENDENTLY VERIFIED · ◆ AA — AGGREGATOR · ◆ VA — VENDOR · ◇ SA — SOURCE-ATTRIBUTED

Variants of this boardSCORING CONFIGS & SPLITS — NOT CITED STANDALONE, SO THEY LIVE HERE

Other tracked prints

THIN BOARDSunder 5 published models each — listed, never hidden
PrintLeaderScoreModels
SuperGPQA (EM)DeepSeek-V4-Pro-Base53.93
THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboardsThe frontier boardModels by vendor