Model comparison
Claude Sonnet 4.5 vs Step 3.7 Flash
Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 37.3 on the Noometry Index. Step 3.7 Flash costs 14× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.
Last verified . 4 shared benchmarks.
StepFunSummary
- They share 4 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 2 categories and Step 3.7 Flash in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3.7 Flash leads 42.9 to 32.3.
- Step 3.7 Flash is cheaper at $0.18 / $1.11 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
- Step 3.7 Flash accepts more context: 256K tokens versus 200K.
- Step 3.7 Flash has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4.5 | Step 3.7 Flash | |
|---|---|---|
| Provider | Anthropic | StepFun |
| Noometry Index | 44.1 | 37.3 |
| Released | 2025-09-29 | 2026-05-29 |
| Weights | Proprietary | Open |
| Context window | 200K | 256K |
| Max output | 64K | 256K |
| Input $ / M tokens | $3 | $0.18 |
| Output $ / M tokens | $15 | $1.11 |
| Results tracked | 73 | 5 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4.5 leads
Claude Sonnet 4.5: 47.3 (#61), Step 3.7 Flash: 40.0 (#150)
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| SciCode | 44.7% | 40% |
| ALE-Bench | 796.15 | 694.12 |
| SWE-bench Verified | 71.3% | — |
| SWE-bench Verified (bash only) | 71.4% | — |
| LMArena WebDev | 1393 | — |
| SWE-bench Multilingual | 67% | — |
| GSO | 14.7% | — |
| WeirdML | 47.7% | — |
| LMArena Coding | 1489 | — |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Not comparable
Claude Sonnet 4.5: 38.3 (#32), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| Terminal-Bench | 46.5% | — |
| Berkeley Function Calling Leaderboard | 73.2% | — |
| GDPval | 42.5% | — |
| Remote Labor Index | 2.1% | — |
| τ²-bench Airline | 72% | — |
| τ²-bench Banking | 25.3% | — |
| τ²-bench Retail | 72.4% | — |
| τ²-bench Telecom | 84.9% | — |
| Cybench | 60% | — |
| DeepResearch Bench | 52.6% | — |
| OSWorld | 62.9% | — |
| LMArena Search | 1159 | — |
| METR Time Horizons | 67.4% | — |
| Vending-Bench 2 | 3,839 | — |
Reasoning Claude Sonnet 4.5 leads
Claude Sonnet 4.5: 26.9 (#125), Step 3.7 Flash: 21.6 (#219)
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| NYT Connections (extended) | 37.3% | 39.7% |
| CritPt | 1.1% | 2.3% |
| ARC-AGI-2 | 13.6% | — |
| SimpleBench | 54.3% | — |
| Kagi LLM Benchmark | 57.9% | — |
| ARC-AGI-1 | 63.7% | — |
| Chess Puzzles | 12% | — |
| EnigmaEval | 6% | — |
| EBR-Bench | 2.4% | — |
| LMArena Hard Prompts | 1462 | — |
| Mystery Game Puzzles | 17% | — |
| DTBench | 83.2% | — |
| LMCA | 38.8% | — |
| Epoch Capabilities Index | 146.84 | — |
| ForecastBench | 61.9 | — |
Math Step 3.7 Flash leads
Claude Sonnet 4.5: 32.3 (#216), Step 3.7 Flash: 42.9 (#82)
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 23.9% | — |
| FrontierMath Tier 4 | 2.4% | — |
| MathArena Final-Answer Competitions | — | 68.5% |
| OTIS Mock AIME 2024-2025 | 77.8% | — |
| ProofBench | 19% | — |
| Omni-MATH | 55.3% | — |
| LMArena Math | 1449 | — |
| MATH Level 5 | 97.7% | — |
| FrontierMath (Feb 2025 set) | 15.2% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Not comparable
Claude Sonnet 4.5: 48.4 (#76), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| GPQA Diamond | 82.3% | — |
| Humanity's Last Exam | 13.7% | — |
| SimpleQA Verified | 30.7% | — |
| MMLU-Pro | 86.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 68.6% | — |
| LMArena Expert | 1482 | — |
Multimodal Not comparable
Claude Sonnet 4.5: 34.8 (#89), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| VPCT | 39.8% | — |
| LMArena Document | 1450 | — |
Multilingual Not comparable
Claude Sonnet 4.5: 53.4 (#69), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| LMArena Non-English | 1425 | — |
| LMArena Chinese | 1459 | — |
| LMArena French | 1458 | — |
| LMArena German | 1427 | — |
| LMArena Japanese | 1390 | — |
| LMArena Korean | 1403 | — |
| LMArena Russian | 1437 | — |
| LMArena Spanish | 1457 | — |
Instruction Following Not comparable
Claude Sonnet 4.5: 75.0 (#78), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| IFEval | 85% | — |
| LMArena Instruction Following | 1459 | — |
Long Context Not comparable
Claude Sonnet 4.5: 45.2 (#46), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| LMArena Longer Query | 1476 | — |
Writing & Preference Not comparable
Claude Sonnet 4.5: 66.5 (#34), Step 3.7 Flash: —
| Benchmark | Claude Sonnet 4.5 | Step 3.7 Flash |
|---|---|---|
| LMArena Text | 1439 | — |
| LMArena Creative Writing | 1442 | — |
| EQ-Bench Creative Writing | 1678 | — |
| WildBench | 85.4% | — |
| LMArena Multi-Turn | 1465 | — |
Frequently asked questions
Is Claude Sonnet 4.5 better than Step 3.7 Flash?
Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 37.3 on the Noometry Index. Step 3.7 Flash costs 14× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.
Which is cheaper, Claude Sonnet 4.5 or Step 3.7 Flash?
Step 3.7 Flash is cheaper. It lists at $0.18 per million input tokens and $1.11 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.
Is Claude Sonnet 4.5 or Step 3.7 Flash better for coding?
Claude Sonnet 4.5 scores higher on coding benchmarks: 47.3 versus 40.0 in the Noometry coding category.
Which has the bigger context window?
Step 3.7 Flash does, with 256K tokens against 200K.
How many benchmarks do Claude Sonnet 4.5 and Step 3.7 Flash share?
4 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and Step 3.7 Flash has 5.