Noometry

Model comparison

Claude Opus 4.8 vs Claude Sonnet 4.6

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 50.1 on the Noometry Index. Claude Sonnet 4.6 costs 1.7× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Last verified . 51 shared benchmarks.

Claude Opus 4.8 Anthropic

60.7

Rank #9 Confirmed

Claude Sonnet 4.6 Anthropic

50.1

Rank #42 Confirmed

Summary

  • They share 51 benchmarks with published results for both. Claude Opus 4.8 scores higher in 10 categories and Claude Sonnet 4.6 in 0 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Opus 4.8 leads 78.2 to 51.9.
  • The biggest single-benchmark swing is DeepSWE: 59% for Claude Opus 4.8 and 29.9% for Claude Sonnet 4.6.
  • Claude Sonnet 4.6 is cheaper at $3 / $15 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.

Side by side

Claude Opus 4.8Claude Sonnet 4.6
ProviderAnthropicAnthropic
Noometry Index60.750.1
Released2026-05-282026-02-17
WeightsProprietaryProprietary
Context window1M1M
Max output128K128K
Input $ / M tokens$5$3
Output $ / M tokens$25$15
Results tracked6557

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Claude Opus 4.8 leads

Claude Opus 4.8: 60.2 (#9), Claude Sonnet 4.6: 46.0 (#62)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
DeepSWE59%29.9%
FrontierCode46.5%24.3%
LMArena WebDev15561521
SciCode53.5%46.8%
WeirdML82.9%66.1%
LMArena Coding14911504
ALE-Bench1,5641,327
SWE-bench Verified—75.2%
GSO47.1%—

Agentic & Tool Use Claude Opus 4.8 leads

Claude Opus 4.8: 47.6 (#10), Claude Sonnet 4.6: 39.1 (#26)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
APEX-Agents48.9%43%
OSWorld 2.020.6%9.3%
DeepResearch Bench50.2%54.9%
GBAEval70.9%48.8%
GDP.pdf24%18%
LMArena Search12041221
Vending-Bench 25,7877,204
Terminal-Bench—53.4%
Remote Labor Index8.3%—
τ²-bench Banking39.7%—
OSWorld—72.1%
PostTrainBench33.8%—
ExploitBench—23.6%

Reasoning Claude Opus 4.8 leads

Claude Opus 4.8: 64.6 (#14), Claude Sonnet 4.6: 46.1 (#40)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
ARC-AGI-272.1%60.4%
NYT Connections (extended)91.1%80.9%
ARC-AGI-192.5%86.5%
CritPt20.9%3.1%
Chess Puzzles34%13%
LMArena Hard Prompts14811485
Mystery Game Puzzles36%16%
DTBench94.9%89.9%
LMCA57.5%46.5%
Epoch Capabilities Index158.3152.25
ForecastBench59.962
SimpleBench64.8%—
Kagi LLM Benchmark88.8%—
EnigmaEval23.5%—
Thematic Generalization—76.3%
EBR-Bench28.6%—
Surface Evolver Bench87.5%—
Bench to the Future 30.14—

Math Claude Opus 4.8 leads

Claude Opus 4.8: 78.2 (#11), Claude Sonnet 4.6: 51.9 (#41)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
OTIS Mock AIME 2024-202598.3%85.8%
ProofBench69%45%
LMArena Math14871463
FrontierMath (Feb 2025 set)47.2%32.4%
FrontierMath Tier 4 (v1)31.3%8.3%
FrontierMath (Tiers 1-3)80%—
FrontierMath Tier 456.1%—
MathArena Final-Answer Competitions91.8%—

Knowledge Claude Opus 4.8 leads

Claude Opus 4.8: 61.3 (#26), Claude Sonnet 4.6: 51.6 (#60)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
GPQA Diamond91%87.4%
SimpleQA Verified53%35.5%
LMArena Expert15011500
Vectara Hallucination Rate—10.6%

Multimodal Claude Opus 4.8 leads

Claude Opus 4.8: 43.2 (#20), Claude Sonnet 4.6: 38.0 (#61)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
LMArena Vision12951283
Blueprint-Bench 214.5%6.7%
LMArena Document14751482
Furniture Assembly42.5%—

Multilingual Too close to call

Claude Opus 4.8: 55.4 (#31), Claude Sonnet 4.6: 54.6 (#39)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
LMArena Non-English14501440
LMArena Chinese15091491
LMArena French14821465
LMArena German14701428
LMArena Japanese14401419
LMArena Korean14321411
LMArena Russian14761440
LMArena Spanish14701465

Instruction Following Too close to call

Claude Opus 4.8: 77.6 (#22), Claude Sonnet 4.6: 77.6 (#23)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
LMArena Instruction Following14761476

Long Context Too close to call

Claude Opus 4.8: 45.4 (#37), Claude Sonnet 4.6: 45.3 (#40)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
LMArena Longer Query14821479

Writing & Preference Claude Opus 4.8 leads

Claude Opus 4.8: 72.1 (#17), Claude Sonnet 4.6: 70.4 (#22)

BenchmarkClaude Opus 4.8Claude Sonnet 4.6
LMArena Text14601458
LMArena Creative Writing14531435
EQ-Bench Creative Writing18401810
EQ-Bench 412811207
LMArena Multi-Turn14761464

Frequently asked questions

Is Claude Opus 4.8 better than Claude Sonnet 4.6?

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 50.1 on the Noometry Index. Claude Sonnet 4.6 costs 1.7× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 4.8 or Claude Sonnet 4.6?

Claude Sonnet 4.6 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Opus 4.8 lists at $5 and $25.

Is Claude Opus 4.8 or Claude Sonnet 4.6 better for coding?

Claude Opus 4.8 scores higher on coding benchmarks: 60.2 versus 46.0 in the Noometry coding category.

Which has the bigger context window?

Both accept 1M tokens.

How many benchmarks do Claude Opus 4.8 and Claude Sonnet 4.6 share?

51 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and Claude Sonnet 4.6 has 57.

Related comparisons

Go deeper