Model comparison
GPT-5 Nano vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 33.5 on the Noometry Index.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. GPT-5 Nano scores higher in 1 category and Step 3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Step 3 leads 54.3 to 39.1.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Step 3 | |
|---|---|---|
| Provider | OpenAI | StepFun |
| Noometry Index | 33.5 | 40.5 |
| Released | 2025-08-07 | — |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.05 | — |
| Output $ / M tokens | $0.40 | — |
| Results tracked | 49 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
GPT-5 Nano: 33.6 (#254), Step 3: 40.1 (#147)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Coding | 1351 | 1367 |
| SWE-bench Verified (bash only) | 34.8% | — |
| WeirdML | 38.1% | — |
| ALE-Bench | 718.67 | — |
Agentic & Tool Use Not comparable
GPT-5 Nano: 25.8 (#106), Step 3: —
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| Berkeley Function Calling Leaderboard | 51.5% | — |
Reasoning Step 3 leads
GPT-5 Nano: 16.3 (#306), Step 3: 28.4 (#105)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 62.2% | 62.3% |
| LMArena Hard Prompts | 1328 | 1355 |
| ARC-AGI-2 | 2.6% | — |
| ARC-AGI-1 | 20.7% | — |
| Chess Puzzles | 27% | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 62.7% | — |
| LMCA | 7.9% | — |
| Epoch Capabilities Index | 139.38 | — |
| ForecastBench | 59.1 | — |
Math Step 3 leads
GPT-5 Nano: 29.4 (#241), Step 3: 37.6 (#148)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Math | 1317 | 1366 |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| OTIS Mock AIME 2024-2025 | 81.1% | — |
| ProofBench | 12% | — |
| Omni-MATH | 54.6% | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Too close to call
GPT-5 Nano: 35.9 (#178), Step 3: 36.8 (#164)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Expert | 1321 | 1333 |
| GPQA Diamond | 69.4% | — |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal Step 3 leads
GPT-5 Nano: 31.3 (#108), Step 3: 35.5 (#86)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Vision | 1159 | 1177 |
| VPCT | 37.2% | — |
Multilingual Step 3 leads
GPT-5 Nano: 45.3 (#172), Step 3: 46.3 (#159)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Non-English | 1313 | 1327 |
| LMArena Chinese | 1356 | 1397 |
| LMArena German | 1327 | 1371 |
| LMArena Korean | 1269 | 1269 |
| LMArena Russian | 1296 | 1331 |
| LMArena Spanish | 1360 | 1371 |
| LMArena Japanese | 1226 | — |
Instruction Following GPT-5 Nano leads
GPT-5 Nano: 75.0 (#79), Step 3: 70.4 (#164)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1306 | 1332 |
| IFEval | 93.2% | — |
Long Context Step 3 leads
GPT-5 Nano: 31.3 (#281), Step 3: 40.3 (#157)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Longer Query | 1312 | 1326 |
| Fiction.LiveBench | 44.4% | — |
Writing & Preference Step 3 leads
GPT-5 Nano: 39.1 (#249), Step 3: 54.3 (#151)
| Benchmark | GPT-5 Nano | Step 3 |
|---|---|---|
| LMArena Text | 1320 | 1350 |
| LMArena Creative Writing | 1249 | 1321 |
| LMArena Multi-Turn | 1311 | 1341 |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
Frequently asked questions
Is GPT-5 Nano better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 33.5 on the Noometry Index.
Is GPT-5 Nano or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 33.6 in the Noometry coding category.
How many benchmarks do GPT-5 Nano and Step 3 share?
17 benchmarks have published results for both models. GPT-5 Nano has 49 scored results on Noometry and Step 3 has 17.