Model comparison
Mercury 2 vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 39.1 on the Noometry Index.
Last verified . 11 shared benchmarks.
StepFunSummary
- They share 11 benchmarks with published results for both. Mercury 2 scores higher in 2 categories and Step 3 in 5 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where Step 3 leads 40.1 to 33.5.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| Mercury 2 | Step 3 | |
|---|---|---|
| Provider | Inception | StepFun |
| Noometry Index | 39.1 | 40.5 |
| Released | 2026-02-20 | — |
| Weights | Proprietary | Open |
| Context window | 128K | — |
| Max output | 50K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $0.75 | — |
| Results tracked | 17 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Mercury 2: 33.5 (#255), Step 3: 40.1 (#147)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Coding | 1391 | 1367 |
| LMArena WebDev | 1171 | — |
| SciCode | 38.7% | — |
| WeirdML | 43.2% | — |
| ALE-Bench | 785.58 | — |
Reasoning Step 3 leads
Mercury 2: 23.8 (#170), Step 3: 28.4 (#105)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1362 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| CritPt | 0.8% | — |
Math Not comparable
Mercury 2: —, Step 3: 37.6 (#148)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Math | — | 1366 |
Knowledge Too close to call
Mercury 2: 36.2 (#172), Step 3: 36.8 (#164)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Expert | 1358 | 1333 |
| Vectara Hallucination Rate | 12.3% | — |
Multimodal Not comparable
Mercury 2: —, Step 3: 35.5 (#86)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Too close to call
Mercury 2: 46.6 (#157), Step 3: 46.3 (#159)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Non-English | 1331 | 1327 |
| LMArena Chinese | 1417 | 1397 |
| LMArena Russian | 1304 | 1331 |
| LMArena German | — | 1371 |
| LMArena Korean | — | 1269 |
| LMArena Spanish | — | 1371 |
Instruction Following Too close to call
Mercury 2: 70.2 (#165), Step 3: 70.4 (#164)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1329 | 1332 |
Long Context Too close to call
Mercury 2: 40.5 (#154), Step 3: 40.3 (#157)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Longer Query | 1330 | 1326 |
Writing & Preference Too close to call
Mercury 2: 53.8 (#155), Step 3: 54.3 (#151)
| Benchmark | Mercury 2 | Step 3 |
|---|---|---|
| LMArena Text | 1355 | 1350 |
| LMArena Creative Writing | 1289 | 1321 |
| LMArena Multi-Turn | 1358 | 1341 |
Frequently asked questions
Is Mercury 2 better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 39.1 on the Noometry Index.
Is Mercury 2 or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 33.5 in the Noometry coding category.
How many benchmarks do Mercury 2 and Step 3 share?
11 benchmarks have published results for both models. Mercury 2 has 17 scored results on Noometry and Step 3 has 17.