Coding · Benchmark profile

LiveCodeBench

A continuously updated benchmark for code generation, execution, self-repair and related tasks.

Ranking weightedCurrent / rollingRolling38 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on LiveCodeBench

Leaderboard · 38 models

Gemini 3 Pro leads at 91.7%, followed by Qwen3.7-Max (91.6%) and Qwen3.7-Plus (89.6%).

Ranking · 3.2%CodingCurrent

#1Closed

Gemini 3 Pro

Google

gemini-3-pro

91.7%

Overall #16 · Context 2M tokens

#2Closed

Qwen3.7-Max

Alibaba Cloud

qwen3.7-max

91.6%

Overall #13 · Context 1M tokens

#3Closed

Qwen3.7-Plus

Alibaba Cloud

qwen3.7-plus

89.6%

Overall #14 · Context 1M tokens

Visual analysis

See the field, not just the table.

Three views of the same sourced leaderboard: model placement, the score curve, and descriptive provider averages. These visuals never alter ranking calculations.

ModelsColours

GoogleAlibaba CloudZ.AIOpenAIXiaomiDeepSeekMoonshot AIAnthropicxAIMiniMax

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Gemini 3 Pro	#1	Google	91.7%
Qwen3.7-Max	#2	Alibaba Cloud	91.6%
Qwen3.7-Plus	#3	Alibaba Cloud	89.6%
GLM-4.7	#4	Z.AI	89.4%
GPT-5.2	#5	OpenAI	88.9%
Qwen3.6 Plus	#6	Alibaba Cloud	87.1%
MiMo-V2-Flash	#7	Xiaomi	86.8%
DeepSeek V3.2	#8	DeepSeek	86.2%
o4-mini	#9	OpenAI	85.9%
Kimi K2 Thinking	#10	Moonshot AI	85.3%
Kimi K2.5	#11	Moonshot AI	85%
Claude Opus 4.5	#12	Anthropic	84.8%
GPT-5	#13	OpenAI	84.6%
GPT-5 Codex	#14	OpenAI	84.0%
Grok 4 Fast	#15	xAI	83.2%
MiniMax M2	#16	MiniMax	82.6%
Grok 4	#17	xAI	81.9%
MiniMax M2.1	#18	MiniMax	81.0%
o3	#19	OpenAI	80.8%
Gemini 2.5 Pro	#20	Google	80.1%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Gemini 3 Pro	#1	Google	91.7%
Qwen3.7-Max	#2	Alibaba Cloud	91.6%
Qwen3.7-Plus	#3	Alibaba Cloud	89.6%
GLM-4.7	#4	Z.AI	89.4%
GPT-5.2	#5	OpenAI	88.9%
Qwen3.6 Plus	#6	Alibaba Cloud	87.1%
MiMo-V2-Flash	#7	Xiaomi	86.8%
DeepSeek V3.2	#8	DeepSeek	86.2%
o4-mini	#9	OpenAI	85.9%
Kimi K2 Thinking	#10	Moonshot AI	85.3%
Kimi K2.5	#11	Moonshot AI	85%
Claude Opus 4.5	#12	Anthropic	84.8%
GPT-5	#13	OpenAI	84.6%
GPT-5 Codex	#14	OpenAI	84.0%
Grok 4 Fast	#15	xAI	83.2%
MiniMax M2	#16	MiniMax	82.6%
Grok 4	#17	xAI	81.9%
MiniMax M2.1	#18	MiniMax	81.0%
o3	#19	OpenAI	80.8%
Gemini 2.5 Pro	#20	Google	80.1%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Alibaba Cloud	3	89.4%
Z.AI	1	89.4%
Xiaomi	1	86.8%
DeepSeek	1	86.2%
Google	2	85.9%
Moonshot AI	2	85.1%
OpenAI	5	84.8%
Anthropic	1	84.8%
xAI	2	82.5%
MiniMax	2	81.8%

Benchmark score table (38 models)

One best score per model · higher is better

LiveCodeBench leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Gemini 3 Pro gemini-3-pro	Google	closed	91.7%
#2	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	91.6%
#3	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	89.6%
#4	GLM-4.7 glm-4.7	Z.AI	open	89.4%
#5	GPT-5.2 gpt-5.2	OpenAI	closed	88.9%
#6	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	87.1%
#7	MiMo-V2-Flash mimo-v2-flash	Xiaomi	open	86.8%
#8	DeepSeek V3.2 deepseek-v3.2	DeepSeek	open	86.2%
#9	o4-mini o4-mini	OpenAI	closed	85.9%
#10	Kimi K2 Thinking kimi-k2-thinking	Moonshot AI	open	85.3%
#11	Kimi K2.5 kimi-k2.5	Moonshot AI	—	85%
#12	Claude Opus 4.5 claude-opus-4-5	Anthropic	closed	84.8%
#13	GPT-5 gpt-5	OpenAI	closed	84.6%
#14	GPT-5 Codex gpt-5-codex	OpenAI	closed	84.0%
#15	Grok 4 Fast grok-4-fast	xAI	closed	83.2%
#16	MiniMax M2 minimax-m2	MiniMax	open	82.6%
#17	Grok 4 grok-4	xAI	closed	81.9%
#18	MiniMax M2.1 minimax-m2.1	MiniMax	open	81.0%
#19	o3 o3	OpenAI	closed	80.8%
#20	Gemini 2.5 Pro gemini-2.5-pro	Google	closed	80.1%
#21	DeepSeek V3.1 Terminus deepseek-v3.1-terminus	DeepSeek	open	79.8%
#22	Nova 2 Pro nova-2-pro-preview	Amazon	closed	73.0%
#23	Claude Sonnet 4.5 claude-sonnet-4-5	Anthropic	closed	71.4%
#24	Gemini 2.5 Flash gemini-2.5-flash	Google	closed	71.3%
#25	Grok 3 mini grok-3-mini	xAI	closed	69.6%
#26	GLM-4.6 glm-4.6	Z.AI	open	69.5%
#27	GPT-5 mini gpt-5-mini	OpenAI	closed	69.2%
#28	o1 o1	OpenAI	closed	67.9%
#29	Grok Code Fast 1 grok-code-fast-1	xAI	closed	65.7%
#30	Claude Sonnet 4 claude-sonnet-4	Anthropic	closed	65.5%
#31	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	65.4%
#32	Claude Opus 4 claude-opus-4	Anthropic	closed	63.6%
#33	Kimi K2 0905 kimi-k2-0905	Moonshot AI	open	61.0%
#34	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	56.8%
#35	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	55.2%
#36	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	51.1%
#37	Claude Sonnet 3.7 claude-3-7-sonnet	Anthropic	closed	47.3%
#38	Mistral Large 3 mistral-large-2512	Mistral AI	open	46.5%

The top of this snapshot is led by Gemini 3 Pro at 91.7%; third place is 2.1 points behind. The top-10 spread is 6.4 points.

Compare top models

Gemini 3 Pro vs Qwen3.7-Max Qwen3.7-Max vs Qwen3.7-Plus Qwen3.7-Plus vs GLM-4.7 GLM-4.7 vs GPT-5.2

About LiveCodeBench

Definition and scoring

Organisation: LiveCodeBench team
Category: Coding
Version: Not evaluated
Direction: higher is better
Ranking use: Weighted 3.2%
Contamination risk: Medium

A continuously updated benchmark for code generation, execution, self-repair and related tasks. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.

Open benchmark source ↗

FAQ

Quick answers

What does LiveCodeBench measure?

A continuously updated benchmark for code generation, execution, self-repair and related tasks.

Which model scores highest on LiveCodeBench?

Gemini 3 Pro by Google currently leads with 91.7%.

How many models are evaluated on LiveCodeBench?

38 models in the LuminaBench cohort have a qualifying score on this benchmark.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 3.2% of its capability category. Gemini 3 Pro is overall #16.

SWE-bench Verified

Verified · 21 results · Weighted 3.2%

CRUXEval

Rolling · Not evaluated · Not Weighted

SciCode

Rolling · 92 results · Weighted 3.2%

BigCodeBench

Rolling · Not evaluated · Not Weighted

SWE-Bench Pro

v1 · 33 results · Weighted 3.2%

CursorBench

3.2 · 20 results · Weighted 3.2%

Coding · Benchmark profile

LiveCodeBench

A continuously updated benchmark for code generation, execution, self-repair and related tasks.

Ranking weightedCurrent / rollingRolling38 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on LiveCodeBench

Leaderboard · 38 models

Gemini 3 Pro leads at 91.7%, followed by Qwen3.7-Max (91.6%) and Qwen3.7-Plus (89.6%).

Ranking · 3.2%CodingCurrent

#1Closed

Gemini 3 Pro

Google

gemini-3-pro

91.7%

Overall #16 · Context 2M tokens

#2Closed

Qwen3.7-Max

Alibaba Cloud

qwen3.7-max

91.6%

Overall #13 · Context 1M tokens

#3Closed

Qwen3.7-Plus

Alibaba Cloud

qwen3.7-plus

89.6%

Overall #14 · Context 1M tokens

Visual analysis

See the field, not just the table.

Three views of the same sourced leaderboard: model placement, the score curve, and descriptive provider averages. These visuals never alter ranking calculations.

ModelsColours

GoogleAlibaba CloudZ.AIOpenAIXiaomiDeepSeekMoonshot AIAnthropicxAIMiniMax

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Gemini 3 Pro	#1	Google	91.7%
Qwen3.7-Max	#2	Alibaba Cloud	91.6%
Qwen3.7-Plus	#3	Alibaba Cloud	89.6%
GLM-4.7	#4	Z.AI	89.4%
GPT-5.2	#5	OpenAI	88.9%
Qwen3.6 Plus	#6	Alibaba Cloud	87.1%
MiMo-V2-Flash	#7	Xiaomi	86.8%
DeepSeek V3.2	#8	DeepSeek	86.2%
o4-mini	#9	OpenAI	85.9%
Kimi K2 Thinking	#10	Moonshot AI	85.3%
Kimi K2.5	#11	Moonshot AI	85%
Claude Opus 4.5	#12	Anthropic	84.8%
GPT-5	#13	OpenAI	84.6%
GPT-5 Codex	#14	OpenAI	84.0%
Grok 4 Fast	#15	xAI	83.2%
MiniMax M2	#16	MiniMax	82.6%
Grok 4	#17	xAI	81.9%
MiniMax M2.1	#18	MiniMax	81.0%
o3	#19	OpenAI	80.8%
Gemini 2.5 Pro	#20	Google	80.1%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Gemini 3 Pro	#1	Google	91.7%
Qwen3.7-Max	#2	Alibaba Cloud	91.6%
Qwen3.7-Plus	#3	Alibaba Cloud	89.6%
GLM-4.7	#4	Z.AI	89.4%
GPT-5.2	#5	OpenAI	88.9%
Qwen3.6 Plus	#6	Alibaba Cloud	87.1%
MiMo-V2-Flash	#7	Xiaomi	86.8%
DeepSeek V3.2	#8	DeepSeek	86.2%
o4-mini	#9	OpenAI	85.9%
Kimi K2 Thinking	#10	Moonshot AI	85.3%
Kimi K2.5	#11	Moonshot AI	85%
Claude Opus 4.5	#12	Anthropic	84.8%
GPT-5	#13	OpenAI	84.6%
GPT-5 Codex	#14	OpenAI	84.0%
Grok 4 Fast	#15	xAI	83.2%
MiniMax M2	#16	MiniMax	82.6%
Grok 4	#17	xAI	81.9%
MiniMax M2.1	#18	MiniMax	81.0%
o3	#19	OpenAI	80.8%
Gemini 2.5 Pro	#20	Google	80.1%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Alibaba Cloud	3	89.4%
Z.AI	1	89.4%
Xiaomi	1	86.8%
DeepSeek	1	86.2%
Google	2	85.9%
Moonshot AI	2	85.1%
OpenAI	5	84.8%
Anthropic	1	84.8%
xAI	2	82.5%
MiniMax	2	81.8%

Benchmark score table (38 models)

One best score per model · higher is better

LiveCodeBench leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Gemini 3 Pro gemini-3-pro	Google	closed	91.7%
#2	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	91.6%
#3	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	89.6%
#4	GLM-4.7 glm-4.7	Z.AI	open	89.4%
#5	GPT-5.2 gpt-5.2	OpenAI	closed	88.9%
#6	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	87.1%
#7	MiMo-V2-Flash mimo-v2-flash	Xiaomi	open	86.8%
#8	DeepSeek V3.2 deepseek-v3.2	DeepSeek	open	86.2%
#9	o4-mini o4-mini	OpenAI	closed	85.9%
#10	Kimi K2 Thinking kimi-k2-thinking	Moonshot AI	open	85.3%
#11	Kimi K2.5 kimi-k2.5	Moonshot AI	—	85%
#12	Claude Opus 4.5 claude-opus-4-5	Anthropic	closed	84.8%
#13	GPT-5 gpt-5	OpenAI	closed	84.6%
#14	GPT-5 Codex gpt-5-codex	OpenAI	closed	84.0%
#15	Grok 4 Fast grok-4-fast	xAI	closed	83.2%
#16	MiniMax M2 minimax-m2	MiniMax	open	82.6%
#17	Grok 4 grok-4	xAI	closed	81.9%
#18	MiniMax M2.1 minimax-m2.1	MiniMax	open	81.0%
#19	o3 o3	OpenAI	closed	80.8%
#20	Gemini 2.5 Pro gemini-2.5-pro	Google	closed	80.1%
#21	DeepSeek V3.1 Terminus deepseek-v3.1-terminus	DeepSeek	open	79.8%
#22	Nova 2 Pro nova-2-pro-preview	Amazon	closed	73.0%
#23	Claude Sonnet 4.5 claude-sonnet-4-5	Anthropic	closed	71.4%
#24	Gemini 2.5 Flash gemini-2.5-flash	Google	closed	71.3%
#25	Grok 3 mini grok-3-mini	xAI	closed	69.6%
#26	GLM-4.6 glm-4.6	Z.AI	open	69.5%
#27	GPT-5 mini gpt-5-mini	OpenAI	closed	69.2%
#28	o1 o1	OpenAI	closed	67.9%
#29	Grok Code Fast 1 grok-code-fast-1	xAI	closed	65.7%
#30	Claude Sonnet 4 claude-sonnet-4	Anthropic	closed	65.5%
#31	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	65.4%
#32	Claude Opus 4 claude-opus-4	Anthropic	closed	63.6%
#33	Kimi K2 0905 kimi-k2-0905	Moonshot AI	open	61.0%
#34	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	56.8%
#35	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	55.2%
#36	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	51.1%
#37	Claude Sonnet 3.7 claude-3-7-sonnet	Anthropic	closed	47.3%
#38	Mistral Large 3 mistral-large-2512	Mistral AI	open	46.5%

The top of this snapshot is led by Gemini 3 Pro at 91.7%; third place is 2.1 points behind. The top-10 spread is 6.4 points.

Compare top models

Gemini 3 Pro vs Qwen3.7-Max Qwen3.7-Max vs Qwen3.7-Plus Qwen3.7-Plus vs GLM-4.7 GLM-4.7 vs GPT-5.2

About LiveCodeBench

Definition and scoring

Organisation: LiveCodeBench team
Category: Coding
Version: Not evaluated
Direction: higher is better
Ranking use: Weighted 3.2%
Contamination risk: Medium

Open benchmark source ↗

FAQ

Quick answers

What does LiveCodeBench measure?

A continuously updated benchmark for code generation, execution, self-repair and related tasks.

Which model scores highest on LiveCodeBench?

Gemini 3 Pro by Google currently leads with 91.7%.

How many models are evaluated on LiveCodeBench?

38 models in the LuminaBench cohort have a qualifying score on this benchmark.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 3.2% of its capability category. Gemini 3 Pro is overall #16.

SWE-bench Verified

Verified · 21 results · Weighted 3.2%

CRUXEval

Rolling · Not evaluated · Not Weighted

SciCode

Rolling · 92 results · Weighted 3.2%

BigCodeBench

Rolling · Not evaluated · Not Weighted

SWE-Bench Pro

v1 · 33 results · Weighted 3.2%

CursorBench

3.2 · 20 results · Weighted 3.2%

Leaderboard · 38 models

Gemini 3 Pro

Qwen3.7-Max

Qwen3.7-Plus

See the field, not just the table.

Benchmark score table (38 models)

Compare top models

Definition and scoring

Quick answers

What does LiveCodeBench measure?

Which model scores highest on LiveCodeBench?

How many models are evaluated on LiveCodeBench?

Does this affect overall Lumina rank?

More Coding benchmarks

Leaderboard · 38 models

Gemini 3 Pro

Qwen3.7-Max

Qwen3.7-Plus

See the field, not just the table.

Benchmark score table (38 models)

Compare top models

Definition and scoring

Quick answers

What does LiveCodeBench measure?

Which model scores highest on LiveCodeBench?

How many models are evaluated on LiveCodeBench?

Does this affect overall Lumina rank?

More Coding benchmarks