Coding · Benchmark profile

SWE-Bench Pro

Public single-attempt software engineering tasks drawn from repositories outside the original SWE-bench set.

Ranking weightedv1Active30 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on SWE-Bench Pro

Leaderboard · 30 models

Claude Mythos 5 leads at 80.3%, followed by Claude Fable 5 (80%) and Claude Opus 4.8 (69.2%).

Ranking · 3.2%Codingv1

#1Closed

Claude Mythos 5

Anthropic

claude-mythos-5

80.3%

Overall #1 · Context 1M tokens

#2Closed

Claude Fable 5

Anthropic

claude-fable-5

80%

Overall #2 · Context 1M tokens

#3Closed

Claude Opus 4.8

Anthropic

claude-opus-4-8

69.2%

Overall #5 · 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

AnthropicxAIOpenAIZ.AIMetaAlibaba CloudMiniMaxTencentXiaomi

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Claude Mythos 5	#1	Anthropic	80.3%
Claude Fable 5	#2	Anthropic	80%
Claude Opus 4.8	#3	Anthropic	69.2%
Grok 4.5	#4	xAI	64.7%
GPT-5.6 Sol	#5	OpenAI	64.6%
GPT-5.6 Terra	#6	OpenAI	63.4%
Claude Sonnet 5	#7	Anthropic	63.2%
GPT-5.6 Luna	#8	OpenAI	62.7%
GLM-5.2	#9	Z.AI	62.1%
Muse Spark 1.1	#10	Meta	61.5%
Qwen3.7-Max	#11	Alibaba Cloud	60.6%
MiniMax M3	#12	MiniMax	59%
GPT-5.5	#13	OpenAI	58.6%
GLM-5.1	#14	Z.AI	58.4%
Hy3	#15	Tencent	57.9%
GPT-5.4	#16	OpenAI	57.7%
Qwen3.7-Plus	#17	Alibaba Cloud	57.6%
MiMo-V2.5-Pro	#18	Xiaomi	57.2%
Claude Opus 4.5	#19	Anthropic	57.1%
GPT-5.3-Codex	#20	OpenAI	56.8%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Claude Mythos 5	#1	Anthropic	80.3%
Claude Fable 5	#2	Anthropic	80%
Claude Opus 4.8	#3	Anthropic	69.2%
Grok 4.5	#4	xAI	64.7%
GPT-5.6 Sol	#5	OpenAI	64.6%
GPT-5.6 Terra	#6	OpenAI	63.4%
Claude Sonnet 5	#7	Anthropic	63.2%
GPT-5.6 Luna	#8	OpenAI	62.7%
GLM-5.2	#9	Z.AI	62.1%
Muse Spark 1.1	#10	Meta	61.5%
Qwen3.7-Max	#11	Alibaba Cloud	60.6%
MiniMax M3	#12	MiniMax	59%
GPT-5.5	#13	OpenAI	58.6%
GLM-5.1	#14	Z.AI	58.4%
Hy3	#15	Tencent	57.9%
GPT-5.4	#16	OpenAI	57.7%
Qwen3.7-Plus	#17	Alibaba Cloud	57.6%
MiMo-V2.5-Pro	#18	Xiaomi	57.2%
Claude Opus 4.5	#19	Anthropic	57.1%
GPT-5.3-Codex	#20	OpenAI	56.8%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Anthropic	5	70.0%
xAI	1	64.7%
Meta	1	61.5%
OpenAI	6	60.6%
Z.AI	2	60.3%
Alibaba Cloud	2	59.1%
MiniMax	1	59%
Tencent	1	57.9%
Xiaomi	1	57.2%

Benchmark score table (30 models)

One best score per model · higher is better

SWE-Bench Pro leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Claude Mythos 5 claude-mythos-5	Anthropic	closed	80.3%
#2	Claude Fable 5 claude-fable-5	Anthropic	closed	80%
#3	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	69.2%
#4	Grok 4.5 grok-4.5	xAI	closed	64.7%
#5	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	64.6%
#6	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	63.4%
#7	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	63.2%
#8	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	62.7%
#9	GLM-5.2 glm-5-2	Z.AI	open	62.1%
#10	Muse Spark 1.1 muse-spark-1-1	Meta	closed	61.5%
#11	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	60.6%
#12	MiniMax M3 MiniMax-M3	MiniMax	—	59%
#13	GPT-5.5 gpt-5.5	OpenAI	closed	58.6%
#14	GLM-5.1 glm-5-1	Z.AI	open	58.4%
#15	Hy3 hy3	Tencent	open	57.9%
#16	GPT-5.4 gpt-5.4	OpenAI	closed	57.7%
#17	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	57.6%
#18	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	closed	57.2%
#19	Claude Opus 4.5 claude-opus-4-5	Anthropic	closed	57.1%
#20	GPT-5.3-Codex gpt-5.3-codex	OpenAI	closed	56.8%
#21	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	56.6%
#22	MiniMax M2.7 minimax-m2-7	MiniMax	open	56.2%
#23	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	55.1%
#24	GLM-5 glm-5	Z.AI	open	55.1%
#25	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	54.2%
#26	Claude Opus 4.6 claude-opus-4-6	Anthropic	closed	53.4%
#27	Muse Spark muse-spark	Meta	closed	52.4%
#28	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	52.1%
#29	Kimi K2.5 kimi-k2.5	Moonshot AI	—	50.7%
#30	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	49.1%

The top of this snapshot is led by Claude Mythos 5 at 80.3%; third place is 11.1 points behind. The top-10 spread is 18.8 points.

Compare top models

Claude Mythos 5 vs Claude Fable 5 Claude Fable 5 vs Claude Opus 4.8 Claude Opus 4.8 vs Grok 4.5 Grok 4.5 vs GPT-5.6 Sol

About SWE-Bench Pro

Definition and scoring

Organisation: Scale AI
Category: Coding
Version: v1
Direction: higher is better
Ranking use: Weighted 3.2%
Contamination risk: Unknown

Public single-attempt software engineering tasks drawn from repositories outside the original SWE-bench set. 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 SWE-Bench Pro measure?

Public single-attempt software engineering tasks drawn from repositories outside the original SWE-bench set.

Which model scores highest on SWE-Bench Pro?

Claude Mythos 5 by Anthropic currently leads with 80.3%.

How many models are evaluated on SWE-Bench Pro?

30 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. Claude Mythos 5 is overall #1.

LiveCodeBench

Rolling · 38 results · Weighted 3.2%

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

CursorBench

3.2 · 20 results · Weighted 3.2%

Coding · Benchmark profile

SWE-Bench Pro

Public single-attempt software engineering tasks drawn from repositories outside the original SWE-bench set.

Ranking weightedv1Active30 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on SWE-Bench Pro

Leaderboard · 30 models

Claude Mythos 5 leads at 80.3%, followed by Claude Fable 5 (80%) and Claude Opus 4.8 (69.2%).

Ranking · 3.2%Codingv1

#1Closed

Claude Mythos 5

Anthropic

claude-mythos-5

80.3%

Overall #1 · Context 1M tokens

#2Closed

Claude Fable 5

Anthropic

claude-fable-5

80%

Overall #2 · Context 1M tokens

#3Closed

Claude Opus 4.8

Anthropic

claude-opus-4-8

69.2%

Overall #5 · 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

AnthropicxAIOpenAIZ.AIMetaAlibaba CloudMiniMaxTencentXiaomi

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Claude Mythos 5	#1	Anthropic	80.3%
Claude Fable 5	#2	Anthropic	80%
Claude Opus 4.8	#3	Anthropic	69.2%
Grok 4.5	#4	xAI	64.7%
GPT-5.6 Sol	#5	OpenAI	64.6%
GPT-5.6 Terra	#6	OpenAI	63.4%
Claude Sonnet 5	#7	Anthropic	63.2%
GPT-5.6 Luna	#8	OpenAI	62.7%
GLM-5.2	#9	Z.AI	62.1%
Muse Spark 1.1	#10	Meta	61.5%
Qwen3.7-Max	#11	Alibaba Cloud	60.6%
MiniMax M3	#12	MiniMax	59%
GPT-5.5	#13	OpenAI	58.6%
GLM-5.1	#14	Z.AI	58.4%
Hy3	#15	Tencent	57.9%
GPT-5.4	#16	OpenAI	57.7%
Qwen3.7-Plus	#17	Alibaba Cloud	57.6%
MiMo-V2.5-Pro	#18	Xiaomi	57.2%
Claude Opus 4.5	#19	Anthropic	57.1%
GPT-5.3-Codex	#20	OpenAI	56.8%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Claude Mythos 5	#1	Anthropic	80.3%
Claude Fable 5	#2	Anthropic	80%
Claude Opus 4.8	#3	Anthropic	69.2%
Grok 4.5	#4	xAI	64.7%
GPT-5.6 Sol	#5	OpenAI	64.6%
GPT-5.6 Terra	#6	OpenAI	63.4%
Claude Sonnet 5	#7	Anthropic	63.2%
GPT-5.6 Luna	#8	OpenAI	62.7%
GLM-5.2	#9	Z.AI	62.1%
Muse Spark 1.1	#10	Meta	61.5%
Qwen3.7-Max	#11	Alibaba Cloud	60.6%
MiniMax M3	#12	MiniMax	59%
GPT-5.5	#13	OpenAI	58.6%
GLM-5.1	#14	Z.AI	58.4%
Hy3	#15	Tencent	57.9%
GPT-5.4	#16	OpenAI	57.7%
Qwen3.7-Plus	#17	Alibaba Cloud	57.6%
MiMo-V2.5-Pro	#18	Xiaomi	57.2%
Claude Opus 4.5	#19	Anthropic	57.1%
GPT-5.3-Codex	#20	OpenAI	56.8%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Anthropic	5	70.0%
xAI	1	64.7%
Meta	1	61.5%
OpenAI	6	60.6%
Z.AI	2	60.3%
Alibaba Cloud	2	59.1%
MiniMax	1	59%
Tencent	1	57.9%
Xiaomi	1	57.2%

Benchmark score table (30 models)

One best score per model · higher is better

SWE-Bench Pro leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Claude Mythos 5 claude-mythos-5	Anthropic	closed	80.3%
#2	Claude Fable 5 claude-fable-5	Anthropic	closed	80%
#3	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	69.2%
#4	Grok 4.5 grok-4.5	xAI	closed	64.7%
#5	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	64.6%
#6	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	63.4%
#7	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	63.2%
#8	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	62.7%
#9	GLM-5.2 glm-5-2	Z.AI	open	62.1%
#10	Muse Spark 1.1 muse-spark-1-1	Meta	closed	61.5%
#11	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	60.6%
#12	MiniMax M3 MiniMax-M3	MiniMax	—	59%
#13	GPT-5.5 gpt-5.5	OpenAI	closed	58.6%
#14	GLM-5.1 glm-5-1	Z.AI	open	58.4%
#15	Hy3 hy3	Tencent	open	57.9%
#16	GPT-5.4 gpt-5.4	OpenAI	closed	57.7%
#17	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	57.6%
#18	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	closed	57.2%
#19	Claude Opus 4.5 claude-opus-4-5	Anthropic	closed	57.1%
#20	GPT-5.3-Codex gpt-5.3-codex	OpenAI	closed	56.8%
#21	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	56.6%
#22	MiniMax M2.7 minimax-m2-7	MiniMax	open	56.2%
#23	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	55.1%
#24	GLM-5 glm-5	Z.AI	open	55.1%
#25	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	54.2%
#26	Claude Opus 4.6 claude-opus-4-6	Anthropic	closed	53.4%
#27	Muse Spark muse-spark	Meta	closed	52.4%
#28	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	52.1%
#29	Kimi K2.5 kimi-k2.5	Moonshot AI	—	50.7%
#30	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	49.1%

The top of this snapshot is led by Claude Mythos 5 at 80.3%; third place is 11.1 points behind. The top-10 spread is 18.8 points.

Compare top models

Claude Mythos 5 vs Claude Fable 5 Claude Fable 5 vs Claude Opus 4.8 Claude Opus 4.8 vs Grok 4.5 Grok 4.5 vs GPT-5.6 Sol

About SWE-Bench Pro

Definition and scoring

Organisation: Scale AI
Category: Coding
Version: v1
Direction: higher is better
Ranking use: Weighted 3.2%
Contamination risk: Unknown

Open benchmark source ↗

FAQ

Quick answers

What does SWE-Bench Pro measure?

Public single-attempt software engineering tasks drawn from repositories outside the original SWE-bench set.

Which model scores highest on SWE-Bench Pro?

Claude Mythos 5 by Anthropic currently leads with 80.3%.

How many models are evaluated on SWE-Bench Pro?

30 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. Claude Mythos 5 is overall #1.

LiveCodeBench

Rolling · 38 results · Weighted 3.2%

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

CursorBench

3.2 · 20 results · Weighted 3.2%

Leaderboard · 30 models

Claude Mythos 5

Claude Fable 5

Claude Opus 4.8

See the field, not just the table.

Benchmark score table (30 models)

Compare top models

Definition and scoring

Quick answers

What does SWE-Bench Pro measure?

Which model scores highest on SWE-Bench Pro?

How many models are evaluated on SWE-Bench Pro?

Does this affect overall Lumina rank?

More Coding benchmarks

Leaderboard · 30 models

Claude Mythos 5

Claude Fable 5

Claude Opus 4.8

See the field, not just the table.

Benchmark score table (30 models)

Compare top models

Definition and scoring

Quick answers

What does SWE-Bench Pro measure?

Which model scores highest on SWE-Bench Pro?

How many models are evaluated on SWE-Bench Pro?

Does this affect overall Lumina rank?

More Coding benchmarks