Arc Agi 2

Arc Agi 2 benchmark scores from Epoch AI's benchmark data collection.

aggregateepoch_ai

Strengths

Sourced from Epoch AI's systematic benchmark data collection covering frontier models.

Caveats

Auto-generated definition — editorial description not yet written.

How to interpret scores

Contamination risk: moderateFreshness: periodic

Higher is better.

Caveat to keep in mind

Auto-generated definition — editorial description not yet written.

Leaderboard

Last synced: Apr 7, 2026, 4:00 PM

#	Model	Provider	Score	Price/M
1	Gemma 3 270M	Google	100.0	$0.00
2	Qwen3.5 0.8B (Reasoning)	Alibaba	99.9	$0.02
3	Llama 2 Chat 7B	Meta	99.2	$0.10
4	Reka Edge	Rekaai	60.1	$0.10
5	Reka Edge	Reka	60.1	$0.20
6	PALM-2	Google	60.0	$0.00
7	Reka Flash (Sep '24)	Reka AI	20.5	$0.35
8	Qwen2.5 72B Instruct	Alibaba	19.4	$0.19
9	Hermes 3 - Llama-3.1 70B	Nous Research	15.3	$0.30
10	Hermes 4 - Llama-3.1 70B (Non-reasoning)	Nous Research	9.4	$0.20
11	Hermes 4 - Llama-3.1 70B (Reasoning)	Nous Research	9.4	$0.20
12	Sao10K: Llama 3.1 70B Hanami x1	Sao10k	9.4	$3.00
13	Mistral: Ministral 3 3B 2512	Mistral	9.1	$0.10
14	QwQ 32B	Alibaba	6.3	$0.74
15	Gemma 3 1B Instruct	Google	5.7	$0.00
16	WizardLM-2 8x22B	Microsoft Azure	5.2	$0.62
17	DeepHermes 3 - Mistral 24B Preview (Non-reasoning)	Nous Research	5.1	$0.00
18	DeepSeek LLM 67B Chat (V1)	DeepSeek	4.6	$0.00
19	Meta: Llama 3.1 70B Instruct	Meta	4.3	$0.40
20	AlfredPros: CodeLLaMa 7B Instruct Solidity	Alfredpros	3.9	$0.90
21	Granite 4.0 350M	IBM	3.8	$0.00
22	Mistral: Ministral 3 8B 2512	Mistral	3.7	$0.15
23	Mixtral 8x22B Instruct	Mistral	3.4	$0.00
24	Qwen2.5 Instruct 72B	Alibaba	2.7	$0.00
25	LFM2.5-1.2B-Instruct	Liquid AI	2.6	$0.00
26	Google: Gemma 2 27B	Google	2.2	$0.65
27	Qwen3.5 0.8B (Non-reasoning)	Alibaba	2.1	$0.02
28	Qwen2.5 Coder 32B Instruct	Alibaba	2.1	$0.05
29	Qwen2.5 Coder Instruct 32B	Alibaba	2.1	$0.00
30	o1	OpenAI	1.7	$26.25
31	Ministral 3 3B	Mistral	1.6	$0.10
32	Mistral: Ministral 3 14B 2512	Mistral	1.5	$0.20
33	Llama 2 Chat 70B	Meta	1.5	$0.00
34	Google: Gemma 2 9B	Google	1.4	$0.04
35	Meta: Llama 3.3 70B Instruct (free)	Meta	1.4	$0.00
36	Meta: Llama 3.3 70B Instruct	Meta	1.4	$0.15
37	Meta: Llama 3.2 1B Instruct	Meta	1.2	$0.07
38	Claude Sonnet 4.6 (Non-reasoning, High Effort)	Anthropic	1.2	$6.00
39	LFM2 1.2B	Liquid AI	1.0	$0.00
40	GPT-5.4 Pro (xhigh)	OpenAI	0.8	$67.50
41	Gemini 3.1 Pro Preview	Google	0.8	$4.50
42	Claude Opus 4.6 (Adaptive Reasoning, Max Effort)	Anthropic	0.6	$10.00
43	Claude Opus 4.6 (Non-reasoning, High Effort)	Anthropic	0.6	$10.00
44	LFM2.5-VL-1.6B	Liquid AI	0.6	$0.00
45	Cohere: Command R+ (08-2024)	Cohere	0.6	$4.38
46	Cohere: Command R (08-2024)	Cohere	0.6	$0.26
47	Claude Sonnet 4.6 (Non-reasoning, Low Effort)	Anthropic	0.6	$6.00
48	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)	Anthropic	0.6	$6.00
49	Cohere: Command R7B (12-2024)	Cohere	0.6	$0.07
50	Llama 3.2 Instruct 1B	Meta	0.6	$0.10