Skip to main content

Model comparison

Claude 4 Sonnet vs Claude Sonnet 4.6

Data verified

Head-to-head evidence from 16 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

42.79/100
Margin
22.3pts
winning →
65.07/100
1 category wins0 category wins

Public leaderboard positions: Claude 4 Sonnet #158 (Supported); Claude Sonnet 4.6 #32 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Claude 4 Sonnet and Claude Sonnet 4.6 share 16 comparable benchmark results. 1 of 8 categories are comparable. 0 results are unique to Claude 4 Sonnet; 17 to Claude Sonnet 4.6.

Updated July 20, 2026
Shared results
16
Claude 4 Sonnet only
0
Claude Sonnet 4.6 only
17
Comparable categories
1 / 8

Pick Claude Sonnet 4.6 if you want the stronger benchmark profile. Claude 4 Sonnet only becomes the better choice if coding is the priority.

Confidence note. This is a partial-evidence comparison with 16 shared benchmark results across 6 evidence categories; 1 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Claude Sonnet 4.6 is clearly ahead on the BenchAlign aggregate, 65.07 to 42.79. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Claude 4 Sonnet and Claude Sonnet 4.6
CategoryClaude 4 SonnetΔClaude Sonnet 4.6
CodingClaude 4 Sonnet72.7Margin 3.6Claude Sonnet 4.669.1
AgenticClaude 4 SonnetNot measuredMarginNo overlapClaude Sonnet 4.665.2
KnowledgeClaude 4 SonnetNot measuredMarginNo overlapClaude Sonnet 4.666.0
MathClaude 4 SonnetNot measuredMarginNo overlapClaude Sonnet 4.626.4
MultimodalClaude 4 SonnetNot measuredMarginNo overlapClaude Sonnet 4.677.4

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Claude 4 SonnetB · Claude Sonnet 4.6
  1. SWE-bench Verified

    Coding
    Source ↗
    A 72.7%B 79.6%
    Winner: Claude Sonnet 4.6Δ 6.9
    SWE-bench Verified: Claude 4 Sonnet scored 72.7%; Claude Sonnet 4.6 scored 79.6%. Claude Sonnet 4.6 wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricClaude 4 SonnetClaude Sonnet 4.6Comparison
Input / output priceUSD per 1M tokensClaude 4 Sonnet$3 input / $15 outputClaude Sonnet 4.6$3 input / $15 outputListed prices are equal.
Generation speedtokens per secondClaude 4 Sonnet40 tok/sClaude Sonnet 4.644 tok/sClaude Sonnet 4.6 has the higher measured throughput.
First-answer latencyseconds to first tokenClaude 4 Sonnet1.33 sClaude Sonnet 4.61.48 sClaude 4 Sonnet reaches the first token sooner.
Context windowmaximum listed tokensClaude 4 Sonnet200KClaude Sonnet 4.6200KListed context windows are equal.

Benchmark Deep Dive

Agentic
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
τ²-bench resultsSource 52.3%79.5%Claude Sonnet 4.6 leads
Gert LabsSource 39.66%62.92%Claude Sonnet 4.6 leads
JobBenchSource 18.4%36.9%Claude Sonnet 4.6 leads
Terminal-Bench 2.0Source 59.1%Not comparable
OSWorld-VerifiedSource 72.1%Not comparable
Claw-EvalSource 67.8%Not comparable
CyberGymSource 65.2%Not comparable
OSWorld 2.0Source 8.3%Not comparable
CodingClaude 4 Sonnet wins
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
SWE-bench VerifiedSource 72.7%79.6%Claude Sonnet 4.6 leads
AA-SciCodeSource 37.3%46.9%Claude Sonnet 4.6 leads
SWE-RebenchSource 60.7%Not comparable
React Native EvalsSource 80.6%Not comparable
Vibe Code BenchSource 51.48%Not comparable
cursorBench31Source 48.8%Not comparable
FrontierCode 1.1 MainSource 24.3%Not comparable
Reasoning
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
AA-LCRSource 44.3%57.7%Claude Sonnet 4.6 leads
CritPtSource 1.1%0.9%Claude 4 Sonnet leads
Knowledge
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
Artificial Analysis Intelligence IndexSource 25.5%35.9%Claude Sonnet 4.6 leads
AA-GPQA DiamondSource 68.3%79.9%Claude Sonnet 4.6 leads
AA-HLESource 4.0%13.2%Claude Sonnet 4.6 leads
AA-Omniscience IndexSource -9.2%-2.9%Claude Sonnet 4.6 leads
AA-Omniscience AccuracySource 22.4%38.0%Claude Sonnet 4.6 leads
AA-Omniscience Hallucination RateSource 40.8%65.9%Claude 4 Sonnet leads
GPQASource 89.9%Not comparable
SuperGPQASource 95%Not comparable
MMLU-ProSource 79.2%Not comparable
HLESource 49%Not comparable
Math
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
FrontierMath v2 (Tiers 1-3)Source 32.400%Not comparable
FrontierMath v2 (Tier 4)Source 8.300%Not comparable
Multimodal
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
AA-MMMU-ProSource 62.4%70.6%Claude Sonnet 4.6 leads
Design Arena WebsiteSource 11781317Claude Sonnet 4.6 leads
CharXivSource 77.4%Not comparable
Inst. Following
BenchmarkClaude 4 SonnetClaude Sonnet 4.6Result
AA-IFBenchSource 45.4%41.2%Claude 4 Sonnet leads
Frequently Asked Questions (2)

Which is better, Claude 4 Sonnet or Claude Sonnet 4.6?

Claude Sonnet 4.6 is ahead on BenchLM's BenchAlign leaderboard, 65.07 to 42.79. The biggest single separator in this matchup is SWE-bench Verified, where the scores are 72.7% and 79.6%.

Which is better for coding, Claude 4 Sonnet or Claude Sonnet 4.6?

Claude 4 Sonnet has the edge for coding in this comparison, averaging 72.7 versus 69.1. Inside this category, AA-SciCode is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 20, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.