Model comparison
Claude Sonnet 4.6 vs Step 3.7 Flash
Head-to-head evidence from 17 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Verified leaderboard positions: Claude Sonnet 4.6 #13; Step 3.7 Flash unranked
BenchAlign evidence: Claude Sonnet 4.6 supported; Step 3.7 Flash estimated. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Claude Sonnet 4.6 and Step 3.7 Flash share 17 comparable benchmark results. 2 of 8 categories are comparable. 17 results are unique to Claude Sonnet 4.6; 13 to Step 3.7 Flash.
Updated July 16, 2026- Shared results
- 17
- Claude Sonnet 4.6 only
- 17
- Step 3.7 Flash only
- 13
- Comparable categories
- 2 / 8
Pick Claude Sonnet 4.6 if you want the stronger benchmark profile. Step 3.7 Flash only becomes the better choice if agentic is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 17 shared benchmark results across 6 evidence categories; 2 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 provisional aggregate, 76 to 57. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude Sonnet 4.6's sharpest advantage is in coding, where it averages 69.1 against 56.3. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 59.1% to 59.5%. Step 3.7 Flash does hit back in agentic, so the answer changes if that is the part of the workload you care about most.
Claude Sonnet 4.6 is also the more expensive model on tokens at $3.00 input / $15.00 output per 1M tokens, versus $0.20 input / $1.15 output per 1M tokens for Step 3.7 Flash. That is roughly 13.0x on output cost alone. Step 3.7 Flash is the reasoning model in the pair, while Claude Sonnet 4.6 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Step 3.7 Flash gives you the larger context window at 256K, compared with 200K for Claude Sonnet 4.6.
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 | Claude Sonnet 4.6 | Δ | Step 3.7 Flash |
|---|---|---|---|
| Coding | Claude Sonnet 4.669.1 | Margin← 12.8 | Step 3.7 Flash56.3 |
| Agentic | Claude Sonnet 4.665.2 | Margin→ 1.2 | Step 3.7 Flash66.4 |
| Knowledge | Claude Sonnet 4.666.2 | MarginNo overlap | Step 3.7 FlashNot measured |
| Math | Claude Sonnet 4.626.4 | MarginNo overlap | Step 3.7 FlashNot measured |
| Multimodal | Claude Sonnet 4.677.4 | MarginNo overlap | Step 3.7 FlashNot measured |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
Terminal-Bench 2.0
AgenticA 59.1%B 59.5%Winner: Step 3.7 FlashΔ 0.4Terminal-Bench 2.0: Claude Sonnet 4.6 scored 59.1%; Step 3.7 Flash scored 59.5%. Step 3.7 Flash wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Claude Sonnet 4.6 | Step 3.7 Flash | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Claude Sonnet 4.6$3 input / $15 output | Step 3.7 Flash$0.2 input / $1.15 output | Step 3.7 Flash has the lower combined listed price. |
| Generation speedtokens per second | Claude Sonnet 4.644 tok/s | Step 3.7 FlashNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Claude Sonnet 4.61.48 s | Step 3.7 FlashNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Claude Sonnet 4.6200K | Step 3.7 Flash256K | Step 3.7 Flash lists the larger context window. |
Benchmark Deep Dive
AgenticStep 3.7 Flash wins16 benchmarks
| Benchmark | Claude Sonnet 4.6 | Step 3.7 Flash | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 59.1% | 59.5% | Step 3.7 Flash leads |
| OSWorld-VerifiedSource | 72.1% | — | Not comparable |
| Claw-EvalSource | 67.8% | 67.1% | Claude Sonnet 4.6 leads |
| CyberGymSource | 65.2% | — | Not comparable |
| τ²-bench resultsSource | 79.5% | 98.5% | Step 3.7 Flash leads |
| Gert LabsSource | 62.92% | 51.57% | Claude Sonnet 4.6 leads |
| OSWorld 2.0Source | 8.3% | — | Not comparable |
| JobBenchSource | 36.9% | — | Not comparable |
| BrowseCompSource | — | 75.8% | Not comparable |
| DeepSearchQASource | — | 92.8% | Not comparable |
| GDPval-AASource | — | 25.9% | Not comparable |
| ToolathlonSource | — | 49.5% | Not comparable |
| HLE w/ toolsSource | — | 47.2% | Not comparable |
| AA Agentic IndexSource | — | 21.5% | Not comparable |
| GDPval-AASource | — | 1017 | Not comparable |
| APEX-Agents-AASource | — | 14.8% | Not comparable |
CodingClaude Sonnet 4.6 wins11 benchmarks
| Benchmark | Claude Sonnet 4.6 | Step 3.7 Flash | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 79.6% | — | Not comparable |
| 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 |
| Terminal-Bench HardSource | 46.2% | 35.6% | Claude Sonnet 4.6 leads |
| AA-SciCodeSource | 46.9% | 40.0% | Claude Sonnet 4.6 leads |
| FrontierCodeSource | 24.3% | — | Not comparable |
| SWE-bench ProSource | — | 56.3% | Not comparable |
| Terminal-Bench 2.0Source | — | 59.5% | Not comparable |
| AA Coding IndexSource | — | 39.6% | Not comparable |
Reasoning2 benchmarks
Knowledge10 benchmarks
| Benchmark | Claude Sonnet 4.6 | Step 3.7 Flash | Result |
|---|---|---|---|
| GPQASource | 89.9% | — | Not comparable |
| SuperGPQASource | 95% | — | Not comparable |
| MMLU-ProSource | 79.2% | — | Not comparable |
| HLESource | 49% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 35.9% | 30.3% | Claude Sonnet 4.6 leads |
| AA-GPQA DiamondSource | 79.9% | 80.9% | Step 3.7 Flash leads |
| AA-HLESource | 13.2% | 19.9% | Step 3.7 Flash leads |
| AA-Omniscience IndexSource | -2.9% | -37.5% | Claude Sonnet 4.6 leads |
| AA-Omniscience AccuracySource | 38.0% | 25.4% | Claude Sonnet 4.6 leads |
| AA-Omniscience Hallucination RateSource | 65.9% | 84.4% | Claude Sonnet 4.6 leads |
Math2 benchmarks
Multimodal5 benchmarks
Inst. Following1 benchmarks
| Benchmark | Claude Sonnet 4.6 | Step 3.7 Flash | Result |
|---|---|---|---|
| AA-IFBenchSource | 41.2% | 67.3% | Step 3.7 Flash leads |
Frequently Asked Questions (3)
Which is better, Claude Sonnet 4.6 or Step 3.7 Flash?
Claude Sonnet 4.6 is ahead on BenchLM's provisional leaderboard, 76 to 57. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 59.1% and 59.5%.
Which is better for coding, Claude Sonnet 4.6 or Step 3.7 Flash?
Claude Sonnet 4.6 has the edge for coding in this comparison, averaging 69.1 versus 56.3. Inside this category, Terminal-Bench Hard is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, Claude Sonnet 4.6 or Step 3.7 Flash?
Step 3.7 Flash has the edge for agentic tasks in this comparison, averaging 66.4 versus 65.2. Inside this category, τ²-bench results is the benchmark that creates the most daylight between them.
Related Comparisons
The AI models change fast. We track them for you.
A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.