Model comparison
GPT-4.1 vs Step 3.5 Flash
Step 3.5 Flash is the stronger model overall, scoring 42.3 to 35.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. GPT-4.1 scores higher in 0 categories and Step 3.5 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3.5 Flash leads 42.6 to 22.3.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 256K.
- Step 3.5 Flash has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Step 3.5 Flash | |
|---|---|---|
| Provider | OpenAI | StepFun |
| Noometry Index | 35.9 | 42.3 |
| Released | 2025-04-14 | 2026-01-29 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 256K |
| Max output | 33K | 256K |
| Input $ / M tokens | $2 | $0.10 |
| Output $ / M tokens | $8 | $0.30 |
| Results tracked | 52 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3.5 Flash leads
GPT-4.1: 34.4 (#238), Step 3.5 Flash: 42.4 (#105)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Coding | 1391 | 1436 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| WeirdML | 39% | — |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
Agentic & Tool Use Not comparable
GPT-4.1: 34.7 (#43), Step 3.5 Flash: —
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
Reasoning Step 3.5 Flash leads
GPT-4.1: 11.7 (#339), Step 3.5 Flash: 22.2 (#202)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Hard Prompts | 1384 | 1411 |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 28.4% |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | 2.2% | — |
| DTBench | 68.3% | — |
| LMCA | 25.6% | — |
| Epoch Capabilities Index | 136.78 | — |
| ForecastBench | 61.5 | — |
Math Step 3.5 Flash leads
GPT-4.1: 22.3 (#280), Step 3.5 Flash: 42.6 (#84)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Math | 1370 | 1408 |
| FrontierMath (Tiers 1-3) | 6% | — |
| MathArena Final-Answer Competitions | — | 66.8% |
| OTIS Mock AIME 2024-2025 | 38.3% | — |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Step 3.5 Flash leads
GPT-4.1: 37.1 (#160), Step 3.5 Flash: 39.6 (#132)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Expert | 1364 | 1421 |
| GPQA Diamond | 66.9% | — |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Step 3.5 Flash: —
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual Step 3.5 Flash leads
GPT-4.1: 49.4 (#133), Step 3.5 Flash: 50.5 (#119)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1370 | 1385 |
| LMArena Chinese | 1382 | 1447 |
| LMArena French | 1382 | 1421 |
| LMArena German | 1381 | 1405 |
| LMArena Japanese | 1319 | 1354 |
| LMArena Korean | 1339 | 1352 |
| LMArena Russian | 1377 | 1385 |
| LMArena Spanish | 1376 | 1419 |
Instruction Following Step 3.5 Flash leads
GPT-4.1: 71.3 (#153), Step 3.5 Flash: 73.1 (#124)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1367 | 1385 |
| IFEval | 83.8% | — |
Long Context Step 3.5 Flash leads
GPT-4.1: 40.0 (#163), Step 3.5 Flash: 42.8 (#117)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1385 | 1402 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference Step 3.5 Flash leads
GPT-4.1: 57.6 (#125), Step 3.5 Flash: 58.8 (#113)
| Benchmark | GPT-4.1 | Step 3.5 Flash |
|---|---|---|
| LMArena Text | 1383 | 1403 |
| LMArena Creative Writing | 1363 | 1357 |
| LMArena Multi-Turn | 1398 | 1405 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than Step 3.5 Flash?
Step 3.5 Flash is the stronger model overall, scoring 42.3 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or Step 3.5 Flash?
Step 3.5 Flash is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Step 3.5 Flash better for coding?
Step 3.5 Flash scores higher on coding benchmarks: 42.4 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 256K.
How many benchmarks do GPT-4.1 and Step 3.5 Flash share?
17 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Step 3.5 Flash has 19.