livebench coding python 2026_06_25 Benchmark: Scores and Sources

Published result; benchmark version and evaluation conditions remain in the id and result note.

pillar: coding · weight 1 within pillar · unit: % (higher is better) · official board ↗

Row	Model	Result	Normalized (0–100)	Confidence
1	DeepSeek V4.1 Flash DeepSeek	90%	90.0	69% confidence 69 percent, Medium
2	Qwen3.8 27B Alibaba / Qwen	75%	75.0	69% confidence 69 percent, Medium
3	Claude Opus 5 Anthropic	75%	75.0	100% confidence 100 percent, Full
4	DeepSeek V4 Flash Vision Exp DeepSeek	75%	75.0	21% confidence 21 percent, Low
5	Qwen3.8 Flash Next Alibaba / Qwen	70%	70.0	21% confidence 21 percent, Low
6	Claude Fable 5.1 Anthropic	70%	70.0	100% confidence 100 percent, Full
7	Claude Opus 5.5 Anthropic	70%	70.0	100% confidence 100 percent, Full
8	Claude Fable 5 Anthropic	65%	65.0	100% confidence 100 percent, Full
9	Kimi K3 Moonshot AI	65%	65.0	100% confidence 100 percent, Full
10	GPT-6 Astra OpenAI	65%	65.0	100% confidence 100 percent, Full
11	Qwen3.8 Max Alibaba / Qwen	60%	60.0	85% confidence 85 percent, High
12	Claude Haiku 5.5 Anthropic	60%	60.0	100% confidence 100 percent, Full
13	Claude Sonnet 5 Anthropic	60%	60.0	93% confidence 93 percent, High
14	Gemini 3.7 Flash Google	60%	60.0	100% confidence 100 percent, Full
15	Mistral Large 4 mistral	60%	60.0	32% confidence 32 percent, Low
16	GPT-5.2 Codex OpenAI	60%	60.0	43% confidence 43 percent, Low
17	GPT-6.1 Sol OpenAI	60%	60.0	100% confidence 100 percent, Full
18	Grok 4.5 xAI	60%	60.0	100% confidence 100 percent, Full
19	GLM-5.3 Z.ai	60%	60.0	93% confidence 93 percent, High
20	GLM-5.3-Flash Z.ai	60%	60.0	100% confidence 100 percent, Full
21	Muse Spark 1.1 Meta	55%	55.0	61% confidence 61 percent, Medium
22	Muse Spark 1.3 Meta	55%	55.0	93% confidence 93 percent, High
23	Kimi K2.6 Moonshot AI	55%	55.0	85% confidence 85 percent, High
24	GPT-5.2 OpenAI	55%	55.0	100% confidence 100 percent, Full
25	GPT-5.4 mini OpenAI	55%	55.0	100% confidence 100 percent, Full
26	GPT-5.5 OpenAI	55%	55.0	100% confidence 100 percent, Full
27	GPT-5.6 Sol OpenAI	55%	55.0	100% confidence 100 percent, Full
28	GPT-6 Sol OpenAI	55%	55.0	100% confidence 100 percent, Full
29	Grok 4.6 xAI	55%	55.0	100% confidence 100 percent, Full
30	Grok 4.7 xAI	55%	55.0	100% confidence 100 percent, Full
31	GLM-5.2 Z.ai	55%	55.0	100% confidence 100 percent, Full
32	Claude Opus 4.6 Anthropic	50%	50.0	100% confidence 100 percent, Full
33	Claude Opus 4.8 Anthropic	50%	50.0	100% confidence 100 percent, Full
34	Claude Sonnet 4.6 Anthropic	50%	50.0	100% confidence 100 percent, Full
35	DeepSeek V4 Flash 0731 DeepSeek	50%	50.0	69% confidence 69 percent, Medium
36	DeepSeek V4 Pro DeepSeek	50%	50.0	85% confidence 85 percent, High
37	DeepSeek V4 Pro 0813 DeepSeek	50%	50.0	69% confidence 69 percent, Medium
38	Gemini 3.1 Pro Preview Google	50%	50.0	100% confidence 100 percent, Full
39	Gemini 3.5 Flash Google	50%	50.0	100% confidence 100 percent, Full
40	Gemini 3.5 Flash Lite Google	50%	50.0	100% confidence 100 percent, Full
41	Gemini 3.8 Flash Google	50%	50.0	100% confidence 100 percent, Full
42	Muse Spark 1.2 Meta	50%	50.0	53% confidence 53 percent, Medium
43	Kimi K2.7 Code Moonshot AI	50%	50.0	48% confidence 48 percent, Low
44	GPT-5.4 OpenAI	50%	50.0	100% confidence 100 percent, Full
45	GPT-5.4 nano OpenAI	50%	50.0	100% confidence 100 percent, Full
46	GPT-5.6 Terra OpenAI	50%	50.0	100% confidence 100 percent, Full
47	GPT-6 Luna OpenAI	50%	50.0	100% confidence 100 percent, Full
48	Inkling thinkingmachines	50%	50.0	100% confidence 100 percent, Full
49	Qwen3.6 Plus Alibaba / Qwen	45%	45.0	80% confidence 80 percent, High
50	Qwen3.7 Max Alibaba / Qwen	45%	45.0	53% confidence 53 percent, Medium
51	Claude Opus 4.7 Anthropic	45%	45.0	100% confidence 100 percent, Full
52	Nemotron 3 Ultra 550B A55B nvidia	45%	45.0	100% confidence 100 percent, Full
53	GPT-5.6 Luna OpenAI	45%	45.0	100% confidence 100 percent, Full
54	Grok Build 0.1 xAI	45%	45.0	16% confidence 16 percent, Low
55	Qwen3.6 27B Alibaba / Qwen	40%	40.0	53% confidence 53 percent, Medium
56	Claude Opus 4.5 Anthropic	40%	40.0	100% confidence 100 percent, Full
57	Gemini 3.6 Flash Google	40%	40.0	100% confidence 100 percent, Full
58	Claude Sonnet 5.5 Anthropic	35%	35.0	93% confidence 93 percent, High
59	DeepSeek V4 Flash DeepSeek	35%	35.0	53% confidence 53 percent, Medium
60	MiniMax-M3 minimax	35%	35.0	100% confidence 100 percent, Full
61	Grok 4.3 xAI	15%	15.0	80% confidence 80 percent, High

Results are as published by the source behind each value (hover or tap the number). Benchmark scores are shown individually for every source/variant (a model may have multiple rows), and vary by version, harness and date; the normalized column uses the method’s fixed 0–100 scales and feeds the coding pillar of the SI Score.