Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

IFBench leaderboard — the family board.

3 TRACKED VARIANTS · 1 FIRST-CLASS BOARD

One family, 3 tracked prints. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA326 MODELS ON THIS BOARDLEADER INKLING AT 83.4

The board — IFBenchTOP 100 OF 326 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen
01InklingThinking Machines83.42026-09-09
02Grok 4.3SpaceXAI83.32026-05-23
03Grok 4.20SpaceXAI82.92026-09-10
04MiniMax M3MiniMax82.92026-06-04
05Qwen3.8 Max PreviewAlibaba82.82026-08-23
06Inkling-SmallThinking Machines82.22026-08-24
07Nemotron 3 Ultra 550B A55BNVIDIA81.72026-09-01
08Qwen3.8 Flash NextAlibaba81.32026-08-27
09Qwen3.7 MaxAlibaba80.52026-08-23
10Nemotron Cascade 2 30B A3BNVIDIA80.42026-05-02
11MiMo V2.5 ProXiaomi79.92026-09-22
12Nova 2.0 Pro PreviewAmazon79.62026-05-02
13Qwen3.8 27BAlibaba79.52026-08-26
14Granite 4.2 8BIBM79.32026-08-31
15DeepSeek-V4-FlashDeepSeek79.22026-10-01
16Qwen3.7 Plus PreviewAlibaba79.12026-08-26
17Qwen3.5 397B A17BAlibaba78.82026-08-24
18Gemini 3 Flash PreviewGoogle78.02026-05-02
19GPT-5.2 CodexOpenAI77.62026-05-02
20Gemini 3.1 Flash Lite PreviewGoogle77.22026-05-02
21Granite 4.2 30BIBM77.22026-08-31
22Gemini 3.1 ProGoogle77.12026-07-16
23Muse GlimmerMeta772026-08-23
24Qwen3.6 Max PreviewAlibaba76.62026-05-02
25Qwen3.5 27BAlibaba76.52026-08-23
26DeepSeek-V4-ProDeepSeek76.52026-10-01
27Gemini 3.5 FlashGoogle76.32026-09-10
28GLM-5.1Z.ai76.32026-05-02
29Qwen3.5 122B A10BAlibaba76.12026-08-23
30Kimi K2.6Moonshot762026-07-16
31Muse SparkMeta75.92026-05-02
32GPT-5.4 nanoOpenAI75.92026-09-10
33GPT-5.5OpenAI75.92026-06-01
34MiniMax M2.7MiniMax75.72026-08-24
35Gemma 4 31BGoogle75.62026-06-01
36GPT-5.2OpenAI75.42026-08-24
37GPT-5 miniOpenAI75.42026-06-21
38GPT-5.3 CodexOpenAI75.42026-05-02
39Qwen3.6 PlusAlibaba75.22026-08-23
40MAI-Code-1-FlashMicrosoft752026-08-23
41Ling-3.0-flashInclusionAI74.52026-09-09
42Granite 4.2 3BIBM74.32026-08-31
43GPT-5 CodexOpenAI74.12026-05-02
44Command A+Cohere742026-08-23
45GPT-5.4OpenAI73.92026-05-02
46Gemma 4 12BGoogle73.52026-06-14
47GLM-5.2Z.ai73.32026-07-16
48GPT-5.4 miniOpenAI73.32026-06-01
49GLM-5-TurboZ.ai73.22026-05-02
50GPT-5OpenAI73.12026-06-21
51GPT-5.1OpenAI72.92026-06-20
52GPT-5.6 SolOpenAI72.72026-08-12
53Nemotron 3 Super 120B A12BNVIDIA72.62026-09-10
54Qwen3.5 35B A3BAlibaba72.52026-08-23
55Gemma 4 26B A4BGoogle72.42026-08-09
56GLM-5Z.ai72.32026-05-02
57MiniMax M2MiniMax72.32026-06-06
58Nemotron 3.5 LightningNVIDIA71.92026-08-26
59MiMo V2 FlashXiaomi71.82026-05-06
60MiniMax M2.5MiniMax71.62026-08-24
61GPT-5.5 InstantOpenAI71.52026-05-31
62o3OpenAI71.42026-05-02
63GPT-5.6 TerraOpenAI71.22026-07-10
64Solar Pro 3Upstage71.22026-05-02
65NVIDIA Nemotron 3 Nano 30B A3BNVIDIA71.12026-09-10
66Mercury 2Inception712026-08-23
67Qwen3 Max (Reasoning)Alibaba70.92026-09-22
68Nova 2.0 LiteAmazon70.72026-06-01
69Gemini 3 ProGoogle70.42026-08-24
70o1OpenAI70.32026-05-02
71Kimi K2.5Moonshot70.22026-07-16
72MiniMax M2.1MiniMax702026-08-24
73GPT-5.1 CodexOpenAI702026-05-02
74Apriel-v1.6-15B-ThinkerServiceNow69.12026-07-03
75Qwen3.6 27BAlibaba69.12026-08-14
76MAI-Thinking-1Microsoft692026-08-23
77Mistral Medium 3.5 128BMistral692026-08-23
78gpt-oss-120bOpenAI69.02026-05-02
79MiMo V2 ProXiaomi68.82026-05-02
80o4-miniOpenAI68.72026-05-02
81Kimi K2 ThinkingMoonshot68.12026-05-02
82GLM-4.7Z.ai67.92026-06-01
83GPT-5.1 Codex miniOpenAI67.92026-05-02
84GPT-5 nanoOpenAI67.62026-06-21
85MiMo V2 OmniXiaomi67.32026-05-02
86Step 3.7 FlashStepFun67.32026-06-02
87MiMo V2.5Xiaomi67.12026-09-24
88o3-miniOpenAI67.12026-05-02
89Qwen3.5 9BAlibaba66.72026-08-23
90Step 3.5 FlashStepFun66.52026-06-18
91Nex-N2-ProNex AGI66.22026-06-26
92Nova 2.0 Omni Reasoning MediumAmazon66.22026-06-01
93gpt-oss-20bOpenAI65.12026-08-09
94K-EXAONELG AI64.72026-05-02
95Qwen3.6 35B A3BAlibaba64.42026-06-01
96DeepSeek-V3.2-SpecialeDeepSeek63.92026-05-02
97Claude Fable 5Anthropic63.52026-10-04
98Nemotron 3 Nano OmniNVIDIA63.22026-05-02
99Kimi K2.7 CodeMoonshot63.12026-06-17
100K2 Think V2MBZUAI62.82026-06-29
◆ IV — INDEPENDENTLY VERIFIED · ◆ AA — AGGREGATOR · ◆ VA — VENDOR · ◇ SA — SOURCE-ATTRIBUTED

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

IFBench (prompt)

VARIANT6 models · best score per model
#ModelVendorScoreVLast seen
01Nemotron 3 Ultra 550B A55BNVIDIA82.32026-09-01
02Granite 4.2 8BIBM79.32026-08-25
03Granite 4.2 30BIBM77.22026-08-25
04Granite 4.2 3BIBM74.32026-08-25
05Nemotron 3 Super 120B A12BNVIDIA73.42026-07-07

IFBench (prompt loose)

VARIANT6 models · best score per model
#ModelVendorScoreVLast seen
01Nemotron 3 Ultra 550B A55BNVIDIA81.72026-06-10
02DeepSeek-V4-ProDeepSeek79.12026-08-24
03Qwen3.5 397B A17BAlibaba78.22026-06-10
04GLM-5.1Z.ai76.62026-06-10
05MiniMax M2.7MiniMax74.62026-06-10
06Kimi K2.6Moonshot73.72026-06-10
THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboardsThe frontier boardModels by vendor