Step 5 Preview
| Released | 2026-09-16 |
|---|---|
| Price per 1M | $1.05 in · $2.99 out |
| Overall score | 58.4 · rank 46 |
| Context window | 1,000,000 tokens |
| Provider | StepFun |
| Weights | Open weights |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 58.4 | 46 | 28 |
| Overall: Coding | 59.1 | 39 | 24 |
| Overall: Reasoning | 58.5 | 54 | 8 |
| Overall: Agentic work | 59.2 | 42 | 21 |
| Overall: Frontend | 56.9 | 49 | 9 |
| Coding | 60.1 | 27 | 5 |
| Agents | 58.9 | 41 | 10 |
| Reasoning | 59.0 | 55 | 6 |
| Math | 57.3 | 52 | 1 |
| Knowledge | 59.6 | 25 | 1 |
| Frontend | 55.6 | 51 | 3 |
| General | 55.3 | 64 | 1 |
| Multimodal | 54.8 | 53 | 1 |
| Value vs price | +9.4 | 32 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 58.6 ±1.1 | 26 |
| High | 57.0 ±5.8 | 2 |
Evidence
Coding 60.1 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Coding | 69.4% | LiveBench | 2026-11-06 |
| Code Migration v1.1 | 47.0% | Vals AI | 2026-11-09 |
| IOI | 38.1% | Vals AI | 2026-10-29 |
| SWE-bench Multimodal | 40.8% | Vals AI | 2026-11-09 |
| SciCode | 58.9% | Epoch AI | 2026-11-11 |
Agents 58.9 ±1.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 4.0 (Vals run) | 31.8% | Vals AI | 2026-11-09 |
| SRE Bench | 0.4% | Vals AI | 2026-11-05 |
| Terminal-Bench-Science | 2.9% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Agentic | 19.4% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 24.8% | LiveBench | 2026-11-06 |
| EMB v1.1 (Excel financial modeling) | 69.6% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Research | 75.5% | LiveBench | 2026-11-06 |
| Tax Agent Bench | 27.3% | Vals AI | 2026-11-09 |
| Finance Agent v2.1 | 58.9% | Vals AI | 2026-11-09 |
| HLAB-Vals 1.1 (legal agent) | 12.9% | Vals AI | 2026-11-09 |
Reasoning 59.0 ±2.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Extended NYT Connections | 76.6% | Lech Mazur | 2026-11-10 |
| MysteryMechanism | 18.5% | Vals AI | 2026-11-09 |
| CritPt | 20.9% | Epoch AI | 2026-11-11 |
| BioMysteryBench | 67.4% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Reasoning | 83.8% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Data Analysis | 63.5% | LiveBench | 2026-11-06 |
Math 57.3 ±2.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ProofBench v1.1 | 42.0% | Vals AI | 2026-11-09 |
| ProofBench | 42.0% | Epoch AI | 2026-11-11 |
Knowledge 59.6 ±6.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Legal Research Bench | 40.4% | Vals AI | 2026-11-09 |
Frontend 55.6 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Design | 70.8% | LiveBench | 2026-11-06 |
| Vibe Code Bench v1.1 | 70.4% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,572 | LMArena | 2026-11-09 |
General 55.3 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,448 | LMArena | 2026-11-09 |
Multimodal 54.8 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,245 | LMArena | 2026-11-09 |
Names matched to this model
step-5, step-5-preview