Benchmark suiteLatest source data: Jul 16, 2026Checked: July 20, 2026

Coding model rankings.

A focused coding view using Code Arena and SWE-bench today, with clear source slots for Terminal-Bench and Aider-style coding benchmarks as those feeds are wired.

What These Mean

Ranked models

114

Code Arena or SWE-bench signal

SWE-bench matches

Matched to model rows

Top coding index

Kimi K3

83 coding index

Best SWE-bench

76.8%

Claude Opus 4.5 20251101

Coding formula

Code generation plus real issue resolution.

The coding index blends Code Arena and SWE-bench where available, with output speed as a small tie-breaker. Terminal-Bench and Aider remain visible as planned feed slots until we have a reliable ingestion path.

Coding performance

Higher coding index is better. SWE-bench cells show verified model matches only.

Top 30 of 114

Rank	Model	Index	Code Arena	SWE-bench	Speed	Run mode	Sources
#1	Kimi K3 kimi-k3 Moonshot AIProprietary	83coding index	1,679Code Arena rank #1	Not listedNo confident match	63 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#2	Claude Opus 4 20250514 claude-opus-4-20250514 AnthropicProprietary	81coding index	Not listedCode Arena	67.6%SWE-bench rank #18	Not listedoutput speed	bash-onlymini 1.0.0	Code ArenaSWE-benchArtificial Analysis
#3	Claude Opus 4.8 Thinking claude-opus-4-8-thinking AnthropicProprietary	80coding index	1,562Code Arena rank #5	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#4	Claude Opus 4.7 Thinking claude-opus-4-7-thinking AnthropicProprietary	79coding index	1,558Code Arena rank #7	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#5	Claude Opus 4.6 claude-opus-4-6 AnthropicProprietary	77coding index	1,536Code Arena rank #12	75.6%SWE-bench rank #4	43 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#6	Claude Fable 5 claude-fable-5 AnthropicProprietary	76coding index	1,631Code Arena rank #2	Not listedNo confident match	58 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#7	Gemini 3 Pro gemini-3-pro GoogleProprietary	76coding index	1,439Code Arena rank #35	74.2%SWE-bench rank #6	Not listedoutput speed	bash-onlymini 1.15.0	Code ArenaSWE-benchArtificial Analysis
#8	Claude Sonnet 4 20250514 claude-sonnet-4-20250514 AnthropicProprietary	75coding index	Not listedCode Arena	64.9%SWE-bench rank #23	Not listedoutput speed	bash-onlymini 1.0.0	Code ArenaSWE-benchArtificial Analysis
#9	Seed 2.1 Pro Preview seed-2.1-pro-preview ByteDanceProprietary	75coding index	1,534Code Arena rank #14	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#10	GPT-5.6 Sol xHigh gpt-5.6-sol-xhigh OpenAIProprietary	75coding index	1,618Code Arena rank #3	Not listedNo confident match	60 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#11	Claude Opus 4.5 20251101 claude-opus-4-5-20251101 AnthropicProprietary	73coding index	1,466Code Arena rank #27	76.8%SWE-bench rank #1	52 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#12	Qwen3.7 Max 20260517 qwen3.7-max-20260517 AlibabaProprietary	72coding index	1,516Code Arena rank #17	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#13	GPT-5.2 gpt-5.2 OpenAIProprietary	71coding index	1,406Code Arena rank #45	72.8%SWE-bench rank #9	Not listedoutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#14	GLM 5.2 glm-5.2 (max) Z.aiOpen weights	70coding index	1,587Code Arena rank #4	Not listedNo confident match	81 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#15	GPT-5.5 xHigh gpt-5.5-xhigh (codex-harness) OpenAIProprietary	70coding index	1,504Code Arena rank #19	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#16	Claude Opus 4.5 20251101 Thinking 32K claude-opus-4-5-20251101-thinking-32k AnthropicProprietary	68coding index	1,490Code Arena rank #22	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#17	Gemini 3 Flash gemini-3-flash (thinking-minimal) GoogleProprietary	67coding index	1,387Code Arena rank #54	75.8%SWE-bench rank #2	212 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#18	GLM 5 glm-5 Z.aiOpen weights	66coding index	1,430Code Arena rank #41	72.8%SWE-bench rank #8	52 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#19	Grok 4.5 grok-4.5 xAIProprietary	66coding index	1,558Code Arena rank #6	Not listedNo confident match	73 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#20	Claude Opus 4.7 claude-opus-4-7 AnthropicProprietary	65coding index	1,555Code Arena rank #8	Not listedNo confident match	45 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#21	Minimax M2.5 minimax-m2.5 MiniMaxOpen weights	65coding index	1,382Code Arena rank #58	75.8%SWE-bench rank #3	76 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#22	Kimi K2.5 Thinking kimi-k2.5-thinking Moonshot AIOpen weights	65coding index	1,433Code Arena rank #39	70.8%SWE-bench rank #13	53 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#23	Muse Spark 1.1 muse-spark-1.1 MetaProprietary	64coding index	1,538Code Arena rank #11	Not listedNo confident match	113 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#24	Claude Sonnet 5 High claude-sonnet-5-high AnthropicProprietary	64coding index	1,542Code Arena rank #10	Not listedNo confident match	62 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#25	Claude Opus 4.6 Thinking claude-opus-4-6-thinking AnthropicProprietary	64coding index	1,542Code Arena rank #9	Not listedNo confident match	46 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#26	Claude Opus 4.8 claude-opus-4-8 AnthropicProprietary	63coding index	1,534Code Arena rank #13	Not listedNo confident match	55 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#27	GPT-5.4 High gpt-5.4-high OpenAIProprietary	62coding index	1,457Code Arena rank #30	Not listedNo confident match	Not listedoutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-bench
#28	DeepSeek V3.2 deepseek-v3.2 DeepSeekOpen weights	62coding index	1,332Code Arena rank #73	70%SWE-bench rank #15	Not listedoutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis
#29	GLM 5.1 glm-5.1 Z.aiOpen weights	62coding index	1,526Code Arena rank #15	Not listedNo confident match	62 tok/soutput speed	PlannedTerminal-Bench + Aider feed	Code ArenaSWE-benchArtificial Analysis
#30	Claude Sonnet 4.5 20250929 claude-sonnet-4-5-20250929 AnthropicProprietary	61coding index	1,386Code Arena rank #55	71.4%SWE-bench rank #12	49 tok/soutput speed	bash-onlymini 2.0.0	Code ArenaSWE-benchArtificial Analysis

Kimi K3

kimi-k3

Moonshot AIProprietary

83coding index

Code Arena

1,679Code Arena rank #1

SWE-bench

Not listedNo confident match

Speed

63 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

Claude Opus 4 20250514

claude-opus-4-20250514

AnthropicProprietary

81coding index

Code Arena

Not listedCode Arena

SWE-bench

67.6%SWE-bench rank #18

Speed

Not listedoutput speed

Run mode

bash-onlymini 1.0.0

Code ArenaSWE-benchArtificial Analysis

Claude Opus 4.8 Thinking

claude-opus-4-8-thinking

AnthropicProprietary

80coding index

Code Arena

1,562Code Arena rank #5

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

Claude Opus 4.7 Thinking

claude-opus-4-7-thinking

AnthropicProprietary

79coding index

Code Arena

1,558Code Arena rank #7

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

Claude Opus 4.6

claude-opus-4-6

AnthropicProprietary

77coding index

Code Arena

1,536Code Arena rank #12

SWE-bench

75.6%SWE-bench rank #4

Speed

43 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

Claude Fable 5

claude-fable-5

AnthropicProprietary

76coding index

Code Arena

1,631Code Arena rank #2

SWE-bench

Not listedNo confident match

Speed

58 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

Gemini 3 Pro

gemini-3-pro

GoogleProprietary

76coding index

Code Arena

1,439Code Arena rank #35

SWE-bench

74.2%SWE-bench rank #6

Speed

Not listedoutput speed

Run mode

bash-onlymini 1.15.0

Code ArenaSWE-benchArtificial Analysis

Claude Sonnet 4 20250514

claude-sonnet-4-20250514

AnthropicProprietary

75coding index

Code Arena

Not listedCode Arena

SWE-bench

64.9%SWE-bench rank #23

Speed

Not listedoutput speed

Run mode

bash-onlymini 1.0.0

Code ArenaSWE-benchArtificial Analysis

Seed 2.1 Pro Preview

seed-2.1-pro-preview

ByteDanceProprietary

75coding index

Code Arena

1,534Code Arena rank #14

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

#10

GPT-5.6 Sol xHigh

gpt-5.6-sol-xhigh

OpenAIProprietary

75coding index

Code Arena

1,618Code Arena rank #3

SWE-bench

Not listedNo confident match

Speed

60 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#11

Claude Opus 4.5 20251101

claude-opus-4-5-20251101

AnthropicProprietary

73coding index

Code Arena

1,466Code Arena rank #27

SWE-bench

76.8%SWE-bench rank #1

Speed

52 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#12

Qwen3.7 Max 20260517

qwen3.7-max-20260517

AlibabaProprietary

72coding index

Code Arena

1,516Code Arena rank #17

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

#13

GPT-5.2

gpt-5.2

OpenAIProprietary

71coding index

Code Arena

1,406Code Arena rank #45

SWE-bench

72.8%SWE-bench rank #9

Speed

Not listedoutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#14

GLM 5.2

glm-5.2 (max)

Z.aiOpen weights

70coding index

Code Arena

1,587Code Arena rank #4

SWE-bench

Not listedNo confident match

Speed

81 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#15

GPT-5.5 xHigh

gpt-5.5-xhigh (codex-harness)

OpenAIProprietary

70coding index

Code Arena

1,504Code Arena rank #19

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

#16

Claude Opus 4.5 20251101 Thinking 32K

claude-opus-4-5-20251101-thinking-32k

AnthropicProprietary

68coding index

Code Arena

1,490Code Arena rank #22

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

#17

Gemini 3 Flash

gemini-3-flash (thinking-minimal)

GoogleProprietary

67coding index

Code Arena

1,387Code Arena rank #54

SWE-bench

75.8%SWE-bench rank #2

Speed

212 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#18

GLM 5

glm-5

Z.aiOpen weights

66coding index

Code Arena

1,430Code Arena rank #41

SWE-bench

72.8%SWE-bench rank #8

Speed

52 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#19

Grok 4.5

grok-4.5

xAIProprietary

66coding index

Code Arena

1,558Code Arena rank #6

SWE-bench

Not listedNo confident match

Speed

73 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#20

Claude Opus 4.7

claude-opus-4-7

AnthropicProprietary

65coding index

Code Arena

1,555Code Arena rank #8

SWE-bench

Not listedNo confident match

Speed

45 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#21

Minimax M2.5

minimax-m2.5

MiniMaxOpen weights

65coding index

Code Arena

1,382Code Arena rank #58

SWE-bench

75.8%SWE-bench rank #3

Speed

76 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#22

Kimi K2.5 Thinking

kimi-k2.5-thinking

Moonshot AIOpen weights

65coding index

Code Arena

1,433Code Arena rank #39

SWE-bench

70.8%SWE-bench rank #13

Speed

53 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#23

Muse Spark 1.1

muse-spark-1.1

MetaProprietary

64coding index

Code Arena

1,538Code Arena rank #11

SWE-bench

Not listedNo confident match

Speed

113 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#24

Claude Sonnet 5 High

claude-sonnet-5-high

AnthropicProprietary

64coding index

Code Arena

1,542Code Arena rank #10

SWE-bench

Not listedNo confident match

Speed

62 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#25

Claude Opus 4.6 Thinking

claude-opus-4-6-thinking

AnthropicProprietary

64coding index

Code Arena

1,542Code Arena rank #9

SWE-bench

Not listedNo confident match

Speed

46 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#26

Claude Opus 4.8

claude-opus-4-8

AnthropicProprietary

63coding index

Code Arena

1,534Code Arena rank #13

SWE-bench

Not listedNo confident match

Speed

55 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#27

GPT-5.4 High

gpt-5.4-high

OpenAIProprietary

62coding index

Code Arena

1,457Code Arena rank #30

SWE-bench

Not listedNo confident match

Speed

Not listedoutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-bench

#28

DeepSeek V3.2

deepseek-v3.2

DeepSeekOpen weights

62coding index

Code Arena

1,332Code Arena rank #73

SWE-bench

70%SWE-bench rank #15

Speed

Not listedoutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

#29

GLM 5.1

glm-5.1

Z.aiOpen weights

62coding index

Code Arena

1,526Code Arena rank #15

SWE-bench

Not listedNo confident match

Speed

62 tok/soutput speed

Run mode

PlannedTerminal-Bench + Aider feed

Code ArenaSWE-benchArtificial Analysis

#30

Claude Sonnet 4.5 20250929

claude-sonnet-4-5-20250929

AnthropicProprietary

61coding index

Code Arena

1,386Code Arena rank #55

SWE-bench

71.4%SWE-bench rank #12

Speed

49 tok/soutput speed

Run mode

bash-onlymini 2.0.0

Code ArenaSWE-benchArtificial Analysis

Benchmark guide

What the scores mean.

A quick reading key for comparing coding models without confusing source coverage, preference scores, and real issue-resolution benchmarks.

Higher: coding indexMissing rows are not zero

What does the coding index measure?

The coding index combines Code Arena and SWE-bench signals where available, with output speed used as a small tie-breaker. It is a practical coding comparison, not a guarantee that the model will solve every repository task.

How should I read Code Arena versus SWE-bench?

Code Arena is a public preference-style benchmark for coding outputs, while SWE-bench measures real software issue resolution. Code Arena is broader and more available; SWE-bench is more task-specific and harder to match across model names.

Why does Terminal-Bench show as planned?

Terminal-Bench is relevant for agentic coding workflows, but this page does not mix it into rankings until a reliable source feed is wired. Planned benchmarks are called out so the table does not imply hidden or invented scores.

Why are some SWE-bench cells empty?

Empty SWE-bench cells mean there is no public model row or no confident match to the model name in this table. They should be read as missing source coverage, not as a zero percent result.