💻 Coding League

#	Model	AIM	Năng lực
1	Claude Fable 5 Anthropic	94.7	100	NEW
2	GPT-5 OpenAI	94.7	100	NEW
3	o3-pro OpenAI	91.2	96	NEW
4	o3 OpenAI	87.2	91	NEW
5	Claude Opus 4.8 Anthropic	86.9	91	NEW
6	DeepSeek-V4-Pro-Max DeepSeek	85.9	90	NEW
7	Claude Opus 4.7 Anthropic	85.7	90	NEW
8	DeepSeek-V4-Flash-Max DeepSeek	83.4	87	NEW
9	Gemini 2.5 Pro Preview 05-06 Google	82.4	86	NEW
10	Grok 4 xAI	78.4	81	NEW
11	DeepSeek-V3.2 DeepSeek	78.3	81	NEW
12	Gemini 2.5 Pro Preview 03-25 Google	77.9	81	NEW
13	MiniMax M2 MiniMax	77.8	81	NEW
14	Claude Opus 4.5 Anthropic	77.6	80	NEW
15	Claude Opus 4.6 Anthropic	77.4	80	NEW
16	Gemini 3.1 Pro Google	77.2	80	NEW
17	MiniMax M3 MiniMax	77.1	80	NEW
18	Qwen3.7 Max Alibaba	76.9	80	NEW
19	Claude Opus 4 Anthropic	76.9	80	NEW
20	Kimi K2.6 Moonshot AI	76.7	79	NEW
21	MiniMax M2.5 MiniMax	76.7	79	NEW
22	GPT-5.2 OpenAI	76.5	79	NEW
23	Claude Sonnet 4.6 Anthropic	76.0	78	NEW
24	MiMo-V2.5-Pro Xiaomi	75.1	78	NEW
25	Qwen3.6 Plus Alibaba	75.0	77	NEW
26	Nemotron 3 Super (120B) NVIDIA	74.9	77	NEW
27	Gemini 3 Flash Google	74.0	76	NEW
28	GLM-5 Zhipu AI	73.8	76	NEW
29	o4-mini OpenAI	73.5	76	NEW
30	Gemini 2.5 Pro Preview 06-05 Google	71.8	74	NEW
31	DeepSeek-V3.2-Exp DeepSeek	71.6	73	NEW
32	MiniMax M2.1 MiniMax	69.7	71	NEW
33	Gemini 2.5 Pro Google	69.4	71	NEW
34	DeepSeek-R1-0528 DeepSeek	69.3	71	NEW
35	Gemini 2.5 Flash Google	65.7	67	NEW
36	Qwen3-Coder 480B A35B Instruct Alibaba Cloud/Qwen	65.6	67	NEW
37	o1 OpenAI	65.5	66	NEW
38	Claude Sonnet 4 Anthropic	65.1	66	NEW
39	GLM-4.5 Zhipu AI	61.5	62	NEW
40	Qwen3-235B-A22B-Instruct-2507 Alibaba Cloud/Qwen	60.6	61	NEW
41	Qwen3 235B A22B Alibaba Cloud/Qwen	60.6	61	NEW
42	Nova 2 Lite Amazon	58.5	58	NEW
43	o3-mini OpenAI	55.1	54	NEW
44	DeepSeek-V3.1 DeepSeek	54.0	53	NEW
45	GPT-4.1 OpenAI	50.0	48	NEW
46	Qwen3 32B Alibaba Cloud/Qwen	50.0	48	NEW
47	Kimi K2 Instruct Moonshot AI	47.2	45	NEW
48	Qwen3 30B A3B Alibaba/Qwen	45.0	43	NEW
49	Claude 3.5 Sonnet Anthropic	38.8	36	NEW
50	DeepSeek R1 Distill Llama 70B DeepSeek	36.8	33	NEW
51	DeepSeek R1 Distill Qwen 32B DeepSeek	36.3	33	NEW
52	Qwen2.5 72B Instruct Alibaba/Qwen	33.6	30	NEW
53	GPT-4o OpenAI	31.2	27	NEW
54	Phi 4 Reasoning Microsoft	30.8	26	NEW
55	DeepSeek-V3 DeepSeek	30.3	26	NEW
56	Gemini 2.5 Flash-Lite Google	26.7	22	NEW
57	DeepSeek R1 Zero DeepSeek	24.7	19	NEW
58	DeepSeek-V3 0324 DeepSeek	23.4	18	NEW
59	GPT-4.1 mini OpenAI	21.8	16	NEW
60	Llama 4 Maverick Meta	14.1	7	NEW
61	DeepSeek R1 Distill Llama 8B DeepSeek	8.0	0	NEW
62	GPT-4.1 nano OpenAI	8.0	0	NEW

Cách tính điểm — AIM Index

● Năng lực 65%● Phổ biến (mạng xã hội) 25%● Độ mới 10%

SWE-bench VerifiedAider PolyglotLiveCodeBench

Nguồn dữ liệu thật (2)

llm-stats.com ↗aider.chat ↗

Điểm minh hoạ (seed). Pipeline R3/R4 sẽ thay bằng AIM Index thật từ benchmark + mạng xã hội.