LogoLMSpeed
Back to models

Model compare

Claude Opus 4.8 vs Claude Sonnet 4.6

The readout for Claude Opus 4.8 and Claude Sonnet 4.6, before the detailed comparison sheet.

Model A

Claude Opus 4.8

Anthropic

Leading
vs

Model B

Claude Sonnet 4.6

Anthropic

Contender

Key Takeaways

Weighted outcome: Claude Opus 4.8. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.

Decision read

Claude Opus 4.8

Claude Opus 4.8 has the higher weighted result; Model A / B score 76.6 to 23.4.

Evidence depth

148 data points

Includes 42 benchmark rows, 0 audit samples, and 10 provider examples.

Selection signal

Start with Claude Opus 4.8

The charts below split 52 high-signal samples across speed, scores, and audit health.

Change comparison

Switch either side of this report to compare another model with the same LMSpeed data pipeline.

Select a different model to open a new comparison URL.

Comparison sheet

This report only uses LMSpeed data for Claude Opus 4.8 and Claude Sonnet 4.6: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.

Model compare
Claude Opus 4.8
Claude Sonnet 4.6
Overall leaderLeadingContender
Weighted overall score76.6 pts23.4 pts
Benchmark category leads6 categories1 categories
Operational advantagesAverage speed, First-token latencyCheapest input price, Free providers, Provider coverage
Context window1M tokens1M tokens
Max output128K tokens128K tokens
Modalities

Input

TextImageFile

Output

Text

Input

TextImageFile

Output

Text
Features
Text inputImage inputFile inputText outputTool callingStructured outputsJSON modeReasoning
Text inputImage inputFile inputText outputTool callingStructured outputsJSON modeReasoning

The overall result weights benchmark capability categories at 80% and price, API speed/latency, and availability at 20%. Recent test volume does not affect the winner, and missing benchmark categories are excluded.

Model metadata

Model compare
Claude Opus 4.8
Claude Sonnet 4.6
DeveloperAnthropicAnthropic
ReleasedMay 2026Feb 2026
ParametersNo dataNo data
TokenizerClaudeClaude
Knowledge cutoffNo dataNo data
OpenRouter IDanthropic/claude-opus-4.8anthropic/claude-sonnet-4.6
ReferencesNo dataNo data

When to choose each model

This report only uses LMSpeed data for Claude Opus 4.8 and Claude Sonnet 4.6: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.

Claude Opus 4.8

Claude Opus 4.8 is stronger in benchmark categories (Agents, Coding, Reasoning, Math, Multimodal) and operational dimensions (Average speed, First-token latency).

Claude Sonnet 4.6

Claude Sonnet 4.6 is stronger in benchmark categories (Knowledge) and operational dimensions (Cheapest input price, Free providers, Provider coverage).

Benchmark score comparison

Third-party benchmark profile synced into LMSpeed; only metrics available for both models are shown.

Category performance

Compare benchmark category scores on a 0-100 scale. Select a category to inspect the gap.

Model A coverage
8 / 8
Model B coverage
7 / 8
Shared
7 shared categories

Avg. score

Claude Opus 4.8

58.1

Avg. score

Claude Sonnet 4.6

52.9

Agents

Claude Opus 4.8 leads by 9.9

Claude Opus 4.861.8
Claude Sonnet 4.651.9

Coding

Claude Opus 4.8 leads by 8.0

Claude Opus 4.864
Claude Sonnet 4.656

Reasoning

Claude Opus 4.8 leads by 5.8

Claude Opus 4.855.9
Claude Sonnet 4.650.1

Knowledge

Claude Sonnet 4.6 leads by 8.5

Claude Opus 4.862.1
Claude Sonnet 4.670.6

Math

Claude Opus 4.8 leads by 4.6

Claude Opus 4.857.2
Claude Sonnet 4.652.6

Multilingual

Claude Opus 4.8

Claude Opus 4.855
Claude Sonnet 4.6-

Multimodal

Claude Opus 4.8 leads by 13.6

Claude Opus 4.859
Claude Sonnet 4.645.4

Instruction following

Claude Opus 4.8 leads by 6.2

Claude Opus 4.850
Claude Sonnet 4.643.8

Professional benchmark details

Metric-level scores with benchmark source, rank depth, confidence, error, and evaluation date where available.

Aggregatereported

BenchLM overall score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

73.0

Rank #10/112 · confidence 4 · eval date 2026-05-28

+14.0

Claude Sonnet 4.6

59.0

Rank #55/112 · confidence 4 · eval date 2026-02-01

Pricingverified

Blended price

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

$10.00/M

Rank #169/186 · confidence 4

+$4.00/M

Claude Sonnet 4.6

winner

$6.00/M

Rank #159/186 · confidence 4

Pricingverified

Input price

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

$5.00/M

Rank #168/186 · confidence 4

+$2.00/M

Claude Sonnet 4.6

winner

$3.00/M

Rank #159/186 · confidence 4

Pricingverified

Output price

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

$25.00/M

Rank #169/186 · confidence 4

+$10.00/M

Claude Sonnet 4.6

winner

$15.00/M

Rank #159/186 · confidence 4

Reasoningverified

HLE

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

winner

48.7%

Rank #7/216 · confidence 4

+15.1%

Claude Sonnet 4.6

33.6%

Rank #49/216 · confidence 4

Reasoningreported

CritPt

SourceClaude Opus 4.8

Claude Opus 4.8

winner

20.9

Rank #13/88 · confidence 4 · eval date 2026-05-28

+20.0

Claude Sonnet 4.6

0.9

Rank #74/88 · confidence 4 · eval date 2026-02-01

Reasoningverified

GPQA

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

winner

92.0%

Rank #19/218 · confidence 4

+4.5%

Claude Sonnet 4.6

87.5%

Rank #50/218 · confidence 4

Reasoningreported

AA-LCR

SourceClaude Opus 4.8

Claude Opus 4.8

winner

77.7

Rank #37/106 · confidence 4 · eval date 2026-05-28

+9.4

Claude Sonnet 4.6

68.3

Rank #70/106 · confidence 4 · eval date 2026-02-01

Reasoningreported

BenchLM Reasoning score

SourceClaude Opus 4.8

Claude Opus 4.8

55.5

Rank #53/67 · confidence 4 · eval date 2026-05-28

+12.4

Claude Sonnet 4.6

winner

67.9

Rank #41/67 · confidence 4 · eval date 2026-02-01

Codingreported

CursorBench v3.1

SourceClaude Opus 4.8

Claude Opus 4.8

winner

58.4

Rank #3/6 · confidence 4 · eval date 2026-05-28

+9.6

Claude Sonnet 4.6

48.8

Rank #5/6 · confidence 4 · eval date 2026-02-01

Codingreported

EEBench

SourceClaude Opus 4.8

Claude Opus 4.8

winner

51.4

Rank #3/25 · confidence 3 · eval date 2026-05-28

+31.0

Claude Sonnet 4.6

20.4

Rank #14/25 · confidence 3 · eval date 2026-02-01

Codingreported

SWE-bench Verified

SourceClaude Opus 4.8

Claude Opus 4.8

winner

88.6

Rank #3/49 · confidence 4 · eval date 2026-05-28

+9.0

Claude Sonnet 4.6

79.6

Rank #15/49 · confidence 4 · eval date 2026-02-01

Codingreported

FrontierCode 1.1 Main

SourceClaude Opus 4.8

Claude Opus 4.8

winner

46.5

Rank #4/11 · confidence 4 · eval date 2026-05-28

+22.2

Claude Sonnet 4.6

24.3

Rank #11/11 · confidence 4 · eval date 2026-02-01

Codingreported

CADGenBench Generation

SourceClaude Opus 4.8

Claude Opus 4.8

winner

27.4

Rank #6/8 · confidence 3 · eval date 2026-05-28

+5.3

Claude Sonnet 4.6

22.1

Rank #7/8 · confidence 3 · eval date 2026-02-01

Codingreported

BenchLM Coding score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

73.1

Rank #10/87 · confidence 4 · eval date 2026-05-28

+11.3

Claude Sonnet 4.6

61.8

Rank #22/87 · confidence 4 · eval date 2026-02-01

Codingreported

LiveCodeBench (Vals)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

87.8

Rank #10/35 · confidence 4 · eval date 2026-05-28

+5.7

Claude Sonnet 4.6

82.1

Rank #30/35 · confidence 4 · eval date 2026-02-01

Codingreported

SWE-bench (Vals)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

88.6

Rank #10/37 · confidence 4 · eval date 2026-05-28

+11.2

Claude Sonnet 4.6

77.4

Rank #26/37 · confidence 4 · eval date 2026-02-01

Codingverified

SciCode

SourceClaude Opus 4.8 (Adaptive Reasoning, Max Effort)

Claude Opus 4.8

winner

54.4%

Rank #16/89 · confidence 4

+4.3%

Claude Sonnet 4.6

50.1%

Rank #33/89 · confidence 4

Codingreported

AA-SciCode

SourceClaude Opus 4.8

Claude Opus 4.8

winner

54.4

Rank #21/113 · confidence 4 · eval date 2026-05-28

+7.5

Claude Sonnet 4.6

46.9

Rank #49/113 · confidence 4 · eval date 2026-02-01

Mathreported

FrontierMath v2 (Tier 4)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

31.3

Rank #8/36 · confidence 4 · eval date 2026-05-28

+22.9

Claude Sonnet 4.6

8.3

Rank #20/36 · confidence 4 · eval date 2026-02-01

Mathreported

FrontierMath v2 (Tiers 1-3)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

47.2

Rank #8/47 · confidence 4 · eval date 2026-05-28

+14.8

Claude Sonnet 4.6

32.4

Rank #18/47 · confidence 4 · eval date 2026-02-01

Mathreported

BenchLM Math score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

65.4

Rank #23/63 · confidence 4 · eval date 2026-05-28

+16.4

Claude Sonnet 4.6

49.0

Rank #38/63 · confidence 4 · eval date 2026-02-01

Knowledgereported

AA-HLE

SourceClaude Opus 4.8

Claude Opus 4.8

winner

48.7

Rank #6/114 · confidence 4 · eval date 2026-05-28

+35.4

Claude Sonnet 4.6

13.3

Rank #82/114 · confidence 4 · eval date 2026-02-01

Knowledgereported

BenchLM Knowledge score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

86.3

Rank #6/83 · confidence 4 · eval date 2026-05-28

+11.1

Claude Sonnet 4.6

75.2

Rank #27/83 · confidence 4 · eval date 2026-02-01

Knowledgereported

MMLU-Pro (Vals)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

89.6

Rank #7/38 · confidence 4 · eval date 2026-05-28

+2.3

Claude Sonnet 4.6

87.3

Rank #23/38 · confidence 4 · eval date 2026-02-01

Knowledgereported

GPQA Diamond (Vals)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

92.4

Rank #16/37 · confidence 4 · eval date 2026-05-28

+6.8

Claude Sonnet 4.6

85.6

Rank #30/37 · confidence 4 · eval date 2026-02-01

Knowledgereported

AA-Omniscience Accuracy

SourceClaude Opus 4.8

Claude Opus 4.8

winner

48.8

Rank #19/106 · confidence 4 · eval date 2026-05-28

+10.2

Claude Sonnet 4.6

38.6

Rank #40/106 · confidence 4 · eval date 2026-02-01

Knowledgereported

Artificial Analysis Intelligence Index

SourceClaude Opus 4.8

Claude Opus 4.8

winner

42.0

Rank #19/117 · confidence 4 · eval date 2026-05-28

+17.3

Claude Sonnet 4.6

24.7

Rank #68/117 · confidence 4 · eval date 2026-02-01

Knowledgereported

AA-GPQA Diamond

SourceClaude Opus 4.8

Claude Opus 4.8

winner

92.0

Rank #21/113 · confidence 4 · eval date 2026-05-28

+12.1

Claude Sonnet 4.6

79.9

Rank #79/113 · confidence 4 · eval date 2026-02-01

Knowledgereported

AA-Omniscience Hallucination Rate

SourceClaude Opus 4.8

Claude Opus 4.8

39.3

Rank #88/106 · confidence 4 · eval date 2026-05-28

+29.2

Claude Sonnet 4.6

winner

68.5

Rank #57/106 · confidence 4 · eval date 2026-02-01

Multimodalreported

BenchLM Multimodal Grounded score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

87.8

Rank #4/56 · confidence 4 · eval date 2026-05-28

+33.7

Claude Sonnet 4.6

54.1

Rank #44/56 · confidence 4 · eval date 2026-02-01

Multimodalreported

CharXiv

SourceClaude Opus 4.8

Claude Opus 4.8

winner

89.9

Rank #5/30 · confidence 4 · eval date 2026-05-28

+12.5

Claude Sonnet 4.6

77.4

Rank #25/30 · confidence 4 · eval date 2026-02-01

Multimodalreported

Design Arena Website

SourceClaude Opus 4.8

Claude Opus 4.8

1267.0

Rank #31/78 · confidence 4 · eval date 2026-05-28

+30.0

Claude Sonnet 4.6

winner

1297.0

Rank #15/78 · confidence 4 · eval date 2026-02-01

Instruction followingreported

BenchLM Instruction Following score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

75.4

Rank #43/52 · confidence 4 · eval date 2026-05-28

+27.2

Claude Sonnet 4.6

48.2

Rank #50/52 · confidence 4 · eval date 2026-02-01

Instruction followingreported

AA-IFBench

SourceClaude Opus 4.8

Claude Opus 4.8

winner

62.2

Rank #48/84 · confidence 4 · eval date 2026-05-28

+21.0

Claude Sonnet 4.6

41.2

Rank #67/84 · confidence 4 · eval date 2026-02-01

Agentsreported

Gert Labs

SourceClaude Opus 4.8

Claude Opus 4.8

winner

73.0

Rank #1/49 · confidence 4 · eval date 2026-05-28

+10.0

Claude Sonnet 4.6

62.9

Rank #8/49 · confidence 4 · eval date 2026-02-01

Agentsreported

OSWorld-Verified

SourceClaude Opus 4.8

Claude Opus 4.8

winner

83.4

Rank #4/27 · confidence 4 · eval date 2026-05-28

+11.3

Claude Sonnet 4.6

72.1

Rank #16/27 · confidence 4 · eval date 2026-02-01

Agentsreported

OSWorld 2.0

SourceClaude Opus 4.8

Claude Opus 4.8

winner

20.6

Rank #10/19 · confidence 4 · eval date 2026-05-28

+12.3

Claude Sonnet 4.6

8.3

Rank #16/19 · confidence 4 · eval date 2026-02-01

Agentsreported

Terminal-Bench 2.0

SourceClaude Opus 4.8

Claude Opus 4.8

winner

74.6

Rank #14/53 · confidence 4 · eval date 2026-05-28

+15.5

Claude Sonnet 4.6

59.1

Rank #35/53 · confidence 4 · eval date 2026-02-01

Agentsreported

Terminal-Bench 2.1 (Vals)

SourceClaude Opus 4.8

Claude Opus 4.8

winner

71.9

Rank #16/40 · confidence 4 · eval date 2026-05-28

+14.6

Claude Sonnet 4.6

57.3

Rank #26/40 · confidence 4 · eval date 2026-02-01

Agentsreported

BenchLM Agentic score

SourceClaude Opus 4.8

Claude Opus 4.8

winner

76.7

Rank #22/77 · confidence 4 · eval date 2026-05-28

+26.2

Claude Sonnet 4.6

50.5

Rank #55/77 · confidence 4 · eval date 2026-02-01

Agentsreported

τ²-bench results

SourceClaude Opus 4.8

Claude Opus 4.8

winner

94.4

Rank #22/82 · confidence 4 · eval date 2026-05-28

+14.9

Claude Sonnet 4.6

79.5

Rank #55/82 · confidence 4 · eval date 2026-02-01

API audit comparison

Latest completed audits from shared providers, with four safety and integrity score groups plus report links.

Provider
Claude Opus 4.8
Claude Sonnet 4.6
No completed audits are available from shared providers yet.

Provider examples

Speed aggregates and input/output pricing share each provider row for real API selection and migration cost checks.

Provider
Claude Opus 4.8
Claude Sonnet 4.6
UoCode25 tests

Claude Opus 4.8

speed / latency

N/A / N/A

input / output

No data

Claude Sonnet 4.6

speed / latency

32 tok/s / 6074ms

input / output

No data

Tokeness.io15 tests

Claude Opus 4.8

speed / latency

46 tok/s / 2538ms

input / output

No data

Claude Sonnet 4.6

speed / latency

48 tok/s / 1587ms

input / output

No data

FreeModel10 tests

Claude Opus 4.8

speed / latency

882 tok/s / 127ms

input / output

No data

Claude Sonnet 4.6

speed / latency

N/A / N/A

input / output

No data

Fucheers10 tests

Claude Opus 4.8

speed / latency

47 tok/s / 3308ms

input / output

No data

Claude Sonnet 4.6

speed / latency

48 tok/s / 718ms

input / output

No data

Fusecode10 tests

Claude Opus 4.8

speed / latency

43 tok/s / 2368ms

input / output

No data

Claude Sonnet 4.6

speed / latency

N/A / N/A

input / output

No data

兔子API

Claude Opus 4.8

claude-opus-4-8

speed / latency

No data

input / output

$0/request

Claude Sonnet 4.6

claude-sonnet-4-6

speed / latency

No data

input / output

$0/request

熊猫 API

Claude Opus 4.8

熊猫-特供-按量-满血逆-claude-opus-4-8

speed / latency

No data

input / output

$0.015/M/$0.076/M

Claude Sonnet 4.6

熊猫-特供-按量-满血逆-claude-sonnet-4.6

speed / latency

No data

input / output

$0.0078/M/$0.039/M

KFCV50

Claude Opus 4.8

claude-opus-4-8

speed / latency

No data

input / output

$0.017/request

Claude Sonnet 4.6

claude-sonnet-4-6

speed / latency

No data

input / output

$0.010/request

PoloAPI

Claude Opus 4.8

claude-opus-4-8-c

speed / latency

No data

input / output

$0.027/request

Claude Sonnet 4.6

claude-sonnet-4-6-c

speed / latency

No data

input / output

$0.027/request

Crond

Claude Opus 4.8

claude-opus-4-8[次]

speed / latency

No data

input / output

$0.030/request

Claude Sonnet 4.6

claude-sonnet-4-6[次]

speed / latency

No data

input / output

$0.030/request

FAQ

Weighted outcome: Claude Opus 4.8. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.

Why is this comparison indexable?
It has 6 verifiable comparison points, and both models have pricing or benchmark data.
Are missing metrics invented?
No. Metrics without LMSpeed data are omitted from this report.

Rankings are based on community-submitted tests and periodic health probes. Advisory only, not official data.