Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

BBH leaderboard — the family board.

3 TRACKED VARIANTS · 2 FIRST-CLASS BOARDS

One family, 3 tracked prints: BBH (HF Open LLM) is cited standalone and keeps live board of its own. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA62 MODELS ON THIS BOARDLEADER CLAUDE 3.5 SONNET AT 93.1

First-class boards in this familyCITED STANDALONE · EACH KEEPS ITS OWN LIVE BOARD

#BoardModels scored
01BBHTHIS PAGE62

The board — BBHTOP 62 OF 62 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen
01Claude 3.5 SonnetAnthropic93.12026-08-23
02Hunyuan Large InstructTencent89.52026-05-31
03Gemini 1.5 ProGoogle89.22026-08-23
04Qwen3 235B A22BAlibaba88.92026-08-23
05DeepSeek-V3.2-Exp-BaseDeepSeek88.72026-06-13
06Kimi K2 BaseMoonshot88.72026-06-13
07MiMo V2 Flash BaseXiaomi88.52026-06-13
08MiMo V2.5 Pro BaseXiaomi88.42026-06-05
09MiMo V2.5 ProXiaomi88.42026-08-23
10DeepSeek-R1-Distill-Qwen-32BDeepSeek88.32026-06-05
11Step3.5 Flash BaseStepFun88.22026-06-05
12DeepSeek-V3.1-BaseDeepSeek88.22026-06-13
13Gemma 3 27BGoogle87.62026-08-23
14DeepSeek-V3DeepSeek87.52026-05-01
15DeepSeek-V4-Pro-BaseDeepSeek87.52026-06-05
16MiMo V2.5 BaseXiaomi87.22026-06-05
17Nova ProAmazon86.92026-09-01
18DeepSeek-V4-Flash-BaseDeepSeek86.92026-06-05
19Claude 3 OpusAnthropic86.82026-08-23
20Hunyuan LargeTencent86.32026-05-31
21GLM-4.5-BaseZ.ai86.22026-06-05
22Llama 3.1 Instruct 405BMeta85.92026-05-31
23SynLogic 32BMiniMax85.82026-06-05
24Gemma 3 12BGoogle85.72026-08-23
25Qwen2.5 Instruct 32BAlibaba84.52026-06-05
26DeepSeek-V2.5DeepSeek84.32026-08-23
27Granite 4.1 30BIBM83.72026-06-15
28Claude 3 SonnetAnthropic82.92026-08-23
29Qwen2 Instruct (72B)Alibaba82.42026-08-23
30Nova LiteAmazon82.42026-09-01
31Qwen2 72BAlibaba82.42026-09-10
32Llama 3.1 70B InstructMeta81.62026-05-31
33Llama 3 (70B)Meta812026-09-10
34Granite 4.1 30B BaseIBM80.72026-06-15
35Granite 4.1 8BIBM80.52026-06-15
36Qwen2.5 Instruct 72BAlibaba79.82026-05-01
37Nova MicroAmazon79.52026-09-01
38Phi 3.5 MoE InstructMicrosoft79.12026-08-23
39DeepSeek-V2DeepSeek78.92026-09-10
40Granite 4.1 3BIBM75.82026-06-15
41Granite 4.1 8B BaseIBM73.82026-06-15
42Claude 3 HaikuAnthropic73.72026-08-23
43Gemma 3 4BGoogle72.22026-08-23
44Phi 4 Mini InstructMicrosoft70.42026-08-23
45Qwen2.5 7B BaseAlibaba70.42026-05-03
46Granite 3.3 8B InstructIBM69.12026-08-23
47Phi 3.5 Mini InstructMicrosoft692026-08-23
48SynLogic 7BMiniMax66.52026-06-05
49MoonlightMoonshot65.22026-06-12
50Granite 4.1 3B BaseIBM63.82026-06-15
51Qwen2.5 Omni 3BAlibaba61.32026-08-09
52Qwen2.5 3B BaseAlibaba59.52026-05-03
53Llama 3 Instruct 8BMeta57.72026-08-26
54Qwen2.5 3BAlibaba56.32026-06-05
55Mistral 7BMistral56.12026-08-26
56Gemma (7B)Google55.12026-08-26
57Qwen3 4B Base DenseAlibaba54.52026-05-03
58Llama 3.2 3B InstructMeta46.82026-06-05
59Gemma 3n E2B InstructGoogle44.32026-09-10
60Gemma 3 1B InstructGoogle39.12026-08-23
61ERNIE 4.5Baidu30.42026-08-23
62Gemma 3 1B Base DenseGoogle26.32026-05-03
◆ IV — INDEPENDENTLY VERIFIED · ◆ AA — AGGREGATOR · ◆ VA — VENDOR · ◇ SA — SOURCE-ATTRIBUTED

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

BBH (EM)

VARIANT7 models · best score per model
#ModelVendorScoreVLast seen
01DeepSeek-V3.2-BaseDeepSeek87.62026-06-27
02DeepSeek-V3DeepSeek87.52026-07-06
03DeepSeek-V4-Pro-BaseDeepSeek87.52026-09-10
04DeepSeek-V4-Flash-BaseDeepSeek86.92026-09-10
05Llama 3.1 Instruct 405BMeta82.92026-07-06
06Qwen2.5 Instruct 72BAlibaba79.82026-07-06
07DeepSeek-V2DeepSeek78.82026-07-06
THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboardsThe frontier boardModels by vendor