Model comparison
Gemini 2.5 Flash vs Step 3
Gemini 2.5 Flash and Step 3 score almost the same on the Noometry Index (39.0 vs 39.9), so choose on price, context window or the category you care about most.
Last verified . 17 shared benchmarks.
Google
StepFunSummary
- They share 17 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 5 categories and Step 3 in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 3 leads 27.6 to 18.2.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 56.8% for Gemini 2.5 Flash and 62.3% for Step 3.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| Gemini 2.5 Flash | Step 3 | |
|---|---|---|
| Provider | StepFun | |
| Noometry Index | 39.0 | 39.9 |
| Released | 2025-04-17 | — |
| Weights | Proprietary | Open |
| Context window | 1.05M | — |
| Max output | 66K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $2.50 | — |
| Results tracked | 54 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Gemini 2.5 Flash: 35.5 (#210), Step 3: 39.5 (#144)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Coding | 1424 | 1367 |
| SWE-bench Verified (bash only) | 28.7% | — |
| Aider Polyglot | 55.1% | — |
| WeirdML | 41.9% | — |
| ALE-Bench | 661.88 | — |
Agentic & Tool Use Not comparable
Gemini 2.5 Flash: 30.6 (#67), Step 3: —
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| Terminal-Bench | 17.1% | — |
| Berkeley Function Calling Leaderboard | 56.2% | — |
| TheAgentCompany | 41.1% | — |
| BALROG | 33.5% | — |
| Vending-Bench 2 | 548.84 | — |
Reasoning Step 3 leads
Gemini 2.5 Flash: 18.2 (#271), Step 3: 27.6 (#102)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 56.8% | 62.3% |
| LMArena Hard Prompts | 1422 | 1355 |
| ARC-AGI-2 | 2.5% | — |
| SimpleBench | 41.2% | — |
| ARC-AGI-1 | 33.3% | — |
| CritPt | 1.1% | — |
| EnigmaEval | 2.7% | — |
| DTBench | 76.5% | — |
| LMCA | 27.5% | — |
| Epoch Capabilities Index | 143.04 | — |
| ForecastBench | 60.6 | — |
Math Gemini 2.5 Flash leads
Gemini 2.5 Flash: 38.1 (#87), Step 3: 34.9 (#148)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Math | 1415 | 1366 |
| OTIS Mock AIME 2024-2025 | 73.1% | — |
| Omni-MATH | 38.5% | — |
| FrontierMath (Feb 2025 set) | 4.8% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Too close to call
Gemini 2.5 Flash: 36.1 (#162), Step 3: 36.2 (#161)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Expert | 1426 | 1333 |
| Humanity's Last Exam | 12.1% | — |
| MMLU-Pro | 63.9% | — |
| Confabulations | 16.8% | — |
| Vectara Hallucination Rate | 7.8% | — |
| GPQA (HELM) | 39% | — |
Multimodal Gemini 2.5 Flash leads
Gemini 2.5 Flash: 41.2 (#31), Step 3: 34.9 (#81)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Vision | 1254 | 1176 |
| GeoBench | 76% | — |
| VPCT | 46.2% | — |
| SpatialViz-Bench | 36.9% | — |
Multilingual Gemini 2.5 Flash leads
Gemini 2.5 Flash: 52.5 (#83), Step 3: 46.5 (#154)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Non-English | 1410 | 1327 |
| LMArena Chinese | 1450 | 1396 |
| LMArena German | 1419 | 1371 |
| LMArena Korean | 1385 | 1271 |
| LMArena Russian | 1415 | 1331 |
| LMArena Spanish | 1421 | 1372 |
| LMArena French | 1433 | — |
| LMArena Japanese | 1403 | — |
Instruction Following Gemini 2.5 Flash leads
Gemini 2.5 Flash: 75.9 (#50), Step 3: 70.6 (#159)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1332 |
| IFEval | 89.8% | — |
Long Context Gemini 2.5 Flash leads
Gemini 2.5 Flash: 47.5 (#16), Step 3: 40.3 (#152)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Longer Query | 1419 | 1326 |
| Fiction.LiveBench | 77.8% | — |
Writing & Preference Too close to call
Gemini 2.5 Flash: 54.0 (#153), Step 3: 54.5 (#146)
| Benchmark | Gemini 2.5 Flash | Step 3 |
|---|---|---|
| LMArena Text | 1417 | 1351 |
| LMArena Creative Writing | 1400 | 1322 |
| LMArena Multi-Turn | 1408 | 1340 |
| Short-Story Creative Writing | 76.5% | — |
| EQ-Bench Creative Writing | 1137 | — |
| WildBench | 81.7% | — |
Frequently asked questions
Is Gemini 2.5 Flash better than Step 3?
Gemini 2.5 Flash and Step 3 score almost the same on the Noometry Index (39.0 vs 39.9), so choose on price, context window or the category you care about most.
Is Gemini 2.5 Flash or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 39.5 versus 35.5 in the Noometry coding category.
How many benchmarks do Gemini 2.5 Flash and Step 3 share?
17 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and Step 3 has 17.