Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

bbq leaderboard — the family board.

8 TRACKED VARIANTS · 1 FIRST-CLASS BOARD

One family, 8 tracked prints. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA75 MODELS ON THIS BOARDLEADER CLAUDE OPUS 4 AT 99.3

The board — bbqTOP 75 OF 75 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen
01Claude Opus 4Anthropic99.32026-06-18
02Claude Sonnet 4.5Anthropic98.92026-09-07
03gpt-oss-120bOpenAI98.52026-05-10
04Gemini 3 ProGoogle98.42026-09-07
05Claude Sonnet 4Anthropic97.92026-06-21
06o3OpenAI97.92026-05-10
07GLM-4.5-AirZ.ai97.82026-05-10
08Gemini 2.5 FlashGoogle97.72026-05-10
09GPT-5 nanoOpenAI97.62026-09-07
10o1OpenAI97.32026-05-10
11o1-miniOpenAI96.92026-05-10
12GPT-5OpenAI96.82026-09-07
13gpt-oss-20bOpenAI96.72026-09-07
14Grok 3 Mini BetaSpaceXAI96.72026-05-10
15DeepSeek-V3DeepSeek96.72026-05-10
16Qwen3 235B A22B FP8 ThroughputAlibaba96.62026-05-10
17DeepSeek-R1DeepSeek96.62026-09-21
18Gemini 2.5 ProGoogle96.42026-05-10
19GPT-5 miniOpenAI96.32026-09-07
20Qwen3 235B A22B Instruct 2507Alibaba96.22026-09-07
21Gemini 2.0 Pro Exp 02.05Google95.62026-05-10
22Palmyra-X-004Writer95.52026-05-10
23Llama 3.1 70B InstructMeta95.42026-05-10
24Gemini 2.0 Flash (Reasoning)Google95.42026-06-24
25Qwen2.5 Instruct 72BAlibaba95.42026-05-10
26Qwen2 Instruct (72B)Alibaba95.12026-05-10
27GPT-4oOpenAI95.12026-05-10
28Claude 3.5 SonnetAnthropic94.92026-05-10
29Kimi K2 InstructMoonshot94.92026-05-10
30Gemini 2.5 Flash LiteGoogle94.92026-05-10
31Palmyra X5Writer94.82026-05-10
32Gemini 1.5 Flash (001)Google94.72026-05-10
33Gemini 1.5 ProGoogle94.52026-05-10
34Llama 3.1 Instruct 405BMeta94.52026-08-29
35Palmyra FinWriter94.22026-05-10
36GPT-4 TurboOpenAI94.12026-05-10
37o3-miniOpenAI94.12026-05-10
38Claude 3 OpusAnthropic942026-05-10
39o4-miniOpenAI942026-05-10
40Grok 4SpaceXAI93.72026-05-10
41Grok 3 BetaSpaceXAI93.62026-05-10
42Mistral Small 3 (Non-Reasoning)Mistral93.32026-06-19
44Claude Haiku 4.5Anthropic92.82026-09-07
45GPT-4.1OpenAI92.62026-05-10
46Claude 3.7 SonnetAnthropic92.12026-05-10
47GPT-4.1 miniOpenAI92.12026-05-10
48GPT-4.5 PreviewOpenAI922026-05-10
51Qwen3 Next 80B A3BAlibaba91.72026-09-21
52Llama 3 Instruct 70BMeta912026-05-10
53Qwen2.5 Instruct Turbo (7B)Alibaba90.62026-05-10
54IBM Granite 4.0 SmallIBM902026-09-07
55Claude 3 SonnetAnthropic902026-05-10
56Command R+ (Apr '24)Cohere89.92026-05-10
57GPT-5.1OpenAI88.72026-09-07
58GPT-4o miniOpenAI88.22026-05-10
59GPT-4.1 nanoOpenAI87.52026-05-10
60Llama 4 ScoutMeta87.52026-05-10
61DeepSeek-LLM-67B-Chat (V1)DeepSeek86.22026-05-10
62Granite 3.3 8B InstructIBM842026-05-10
63Palmyra MedWriter80.12026-05-10
64DBRX InstructDatabricks79.22026-05-10
65Llama 3.1 8B InstructMeta78.52026-05-10
66Llama 3 Instruct 8BMeta76.52026-05-10
67Granite 4.0 MicroIBM75.22026-09-07
68Command RCohere72.42026-05-10
69OLMo 2 32B Instruct March 2025AllenAI71.42026-05-10
71Mistral Instruct V0.3 (7B)Mistral66.82026-05-10
72Claude 3 HaikuAnthropic62.52026-05-10
73Mistral V0.1 (7B)Mistral56.52026-05-10
74OLMo 2 7B Instruct November 2024AllenAI52.52026-05-10
◆ 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
BBQ Ambig Accuracy (%)Claude 3.7 Sonnet98.84
BBQ Ambig Bias (%)Claude 3 Sonnet5.04
BBQ Disambig Accuracy (%)Claude 3 Sonnet90.44
BBQ Disambig Bias (%)Claude 3 Sonnet1.224
BBQ Accuracy on Ambiguous Questionso10.963
BBQ Accuracy on Unambiguous Questionso4-mini0.953
BBQ P(not stereotyping | ambiguous question, not unknown)o4-mini0.263
THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboardsThe frontier boardModels by vendor