All models › stepfun
step-3
stepfun · Apache 2.0 license · on Arena since 14 Aug 2025
Download
Arena score1350#208 of 416 in Chat, paid models included
Size321 B18.7 B active per token (mixture of experts)
4-bit download195 GBweights only, Q4_K_M
Max context—
Where to use it for free
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Speed on popular hardware
| Hardware | GPU memory | Fits? | Speed |
|---|---|---|---|
| RTX 3060 12 GB | 12 GB | Does not fit | — |
| RTX 4060 Ti 16 GB | 16 GB | Does not fit | — |
| RTX 3090 | 24 GB | Does not fit | — |
| RTX 4090 | 24 GB | Does not fit | — |
| RTX 5090 | 32 GB | Does not fit | — |
| RX 7900 XTX | 24 GB | Does not fit | — |
| 2 × RTX 3090 | 48 GB | Does not fit | — |
| 4 × RTX 3090 | 96 GB | Does not fit | — |
| Mac M4 Max 128 GB | 96 GB | Does not fit | — |
| Ryzen AI Max+ 395 128 GB | 96 GB | Does not fit | — |
| 8 × H100 SXM | 640 GB | Fits in GPU | 14 tok/s |
Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.
GPU memory needed
| Level | 8K context | 32K context | 128K context | Quality vs full |
|---|---|---|---|---|
| Q3_K_M | 194 GB | 292 GB | 685 GB | +4.7% noticeable loss |
| Q4_K_M | 228 GB | 326 GB | 719 GB | +1.2% very small loss |
| Q5_K_M | 262 GB | 360 GB | 753 GB | +0.4% near-lossless |
| Q6_K | 297 GB | 395 GB | 788 GB | +0.2% near-lossless |
| Q8_0 | 374 GB | 473 GB | 866 GB | ≈ 0% same as full |
| F16 | 675 GB | 774 GB | 1167 GB | ≈ 0% same as full |
Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.
Chat score over time
14001500
MayJunJulAugSeptOct
step-3 1350Best paid model
Arena score from 17 Apr 2026 to 15 Oct 2026.
Vision score over time
11501200125013001350
MayJunJulAugSeptOct
step-3 1176Best paid model
Arena score from 10 Apr 2026 to 9 Oct 2026.
Similar free models in Chat
- glm-4.7-flash 1351
- nvidia-nemotron-3-nano-30b-a3b-bf16 1348
- intellect-3 1356
- Qwen3-Coder-480B-A35B-Instruct 1357
- minimax-m1 1343
- gemma-3-27b-it 1358 · 27.4 B