What does Artificial Analysis · IFBench measure?

Artificial Analysis · IFBench is a knowledge benchmark in the BenchGecko catalog. 37 AI models have been tested on it. Scores range from 21.1 to 82.9 out of 100.

Which model leads on Artificial Analysis · IFBench?

MiniMax M3 from minimax leads Artificial Analysis · IFBench with a score of 82.9. The median score across 37 tested models is 56.3.

Is Artificial Analysis · IFBench saturated?

No · the top score is 82.9 out of 100 (83%). There is still meaningful room for improvement on Artificial Analysis · IFBench.

Does Artificial Analysis · IFBench predict performance on other benchmarks?

Yes · Artificial Analysis · IFBench scores correlate 0.96 with ARC-AGI across 5 shared models. Models that do well on Artificial Analysis · IFBench tend to do well on ARC-AGI.

How often is Artificial Analysis · IFBench data refreshed?

BenchGecko pulls updates daily. New model scores on Artificial Analysis · IFBench appear as soon as they are published by Epoch AI or the model provider.

Benchmark · KnowledgeSettled

Artificial Analysis · IFBench

Name: Artificial Analysis · IFBench Benchmark
Creator: BenchGecko
License: https://creativecommons.org/licenses/by/4.0/

Updated 2026-09-22

Models tested

Top score

82.9

MiniMax M3

Median

56.3

min 21.1

Top-5 spread

σ 2.7

Competitive

The Frontier

Best score over time · one chart, every benchmark

Chart type

Frontier on Artificial Analysis · IFBench rose from 23.5 to 82.9 in 17 months · +59.4 points · latest leader MiniMax M3 from minimax.

Pink dots = frontier records · 7 totalClick to open model page

Full rankings

37 models tested · sorted by score

#	Model	Score	Price
1	MiniMax M3· minimax	82.9	$0.30
2	Nemotron 3 Ultra· NVIDIA	81.4	$0.50
3	Qwen3.7 Plus· Alibaba Qwen	78.0	$0.32
4	Gemini 3.1 Pro Preview· Google DeepMind	77.1	$2.00
5	Gemma 4 31B· Google DeepMind	75.6	$0.09
6	GPT-5.3-Codex· OpenAI	75.4	$1.75
7	Command A+· Cohere	73.9	$0.30
8	Gemma 4 26B A4B · Google DeepMind	72.4	$0.07
9	Nemotron 3 Super· NVIDIA	71.5	$0.08
10	o3· OpenAI	71.4	$2.00
11	GPT-5.6 Terra· OpenAI	71.2	$2.00
12	gpt-oss-120b· OpenAI	69.0	$0.04
13	Mistral Medium 3.5· Mistral AI	68.8	$1.50
14	gpt-oss-20b· OpenAI	65.1	$0.02
15	Qwen3.6 35B A3B· Alibaba Qwen	64.4	$0.15
16	Kimi K2.7 Code· moonshotai	63.1	$0.61
17	Qwen3 Next 80B A3B Instruct· Alibaba Qwen	60.7	$0.09
18	North Mini Code (free)· Cohere	57.6	$0.00
19	Trinity Large Thinking· arcee-ai	56.3	$0.25
20	Qwen3.5 4B· Alibaba	52.0	—
21	Qwen3.5 397B A17B· Alibaba Qwen	51.6	$0.55
22	Qwen3.5-122B-A10B· Alibaba Qwen	50.8	$0.26
23	Mistral Small 4· Mistral AI	48.2	$0.15
24	Llama 4 Maverick· Meta	43.0	$0.19
25	Llama 4 Scout· Meta	39.5	$0.10
26	Qwen3.5-9B· Alibaba Qwen	37.8	$0.10
27	Command A· Cohere	36.5	$2.50
28	N Nanbeige4.1 3B· Nanbeige	35.4	—
29	Qwen3 Coder Next· Alibaba Qwen	35.2	$0.12
30	Qwen3.5 2B· Alibaba	31.5	—
31	Reka Flash 3· rekaai	30.4	$0.10
32	NVIDIA Nemotron Nano 9B V2· NVIDIA	27.6	—
33	Molmo2 8B· allenai	26.9	—
34	Granite 4.0 Micro· ibm-granite	24.8	$0.02
35	Phi 4· Microsoft	23.5	$0.07
36	Qwen3.5 0.8B· Alibaba	21.5	—
37	Phi 4 Mini Instruct· Microsoft	21.1	—