Hugging FaceA 197B, 12B active chat model from StepFun.
- Rating
- 4.0
- Downloads
- 13K
- Likes
- 15
- Parameters
- 197B, 12B active
- Context
- 256K
- Licence
- apache-2.0
Too big for this Mac
Its smallest size needs 126 GB of memory, more than M4 · 16 GB can give it.
Ratings
4.0 out of 5, across people's blind votes on its answers and the tests below.
LM Arena ratings of October 28, 2026 (CC BY 4.0): people vote between two answers without knowing which model wrote which. The bar ends at 1488, the best any open model has. The Arena tests the full model; a small size on a Mac answers a little worse.
MMLU-Pro: graduate-level questions in 14 subjects. GPQA Diamond: questions PhD scientists write for each other. Humanity's Last Exam: questions written to be beyond today's models. SWE-bench Verified: real GitHub issues fixed so the project's tests pass. Terminal-Bench 2.0: jobs finished alone in a terminal. AIME 2026: this year's AIME competition problems. Scores are the ones filed on the model's Hugging Face page or the board's own table, read 2026-11-05; some come from the maker.
Step 3.5 Flash runs well from a Mac with an M2 Ultra and 192 GB of memory. 5 of 59 Mac configurations run it well.
Compare with
The nearest models in size.
- Step 3.7 FlashStepFun · 17K downloadsToo big for this Mac · 57 GB
- Qwen3.8 Flash NextAlibaba Qwen · 9.4M downloadsToo big for this Mac · 66 GB
- Swift 1.5 Qwen3.8 Flash Next GSQ RCOukisai · 771K downloadsToo big for this Mac · 67 GB
- Qwen3.8 Flash Next Gyroagentionai · 749K downloadsToo big for this Mac · 58 GB
Sizes
From ggml-org/Step-3.5-Flash-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.
| Size | Download | Memory at 8K | Get |
|---|---|---|---|
| Q4_K | 119 GB | 119 GBestimated | Open |
Run It Anywhere
Step 3.5 Flash is an open model, so any app that runs GGUF files can run it. These take the Q4_K size from ggml-org straight from Hugging Face.
- LM StudioMac, Windows, Linux
Open - JanMac, Windows, Linux
Open - MstyMac, Windows, Linux
Open - OllamaMac, Windows, Linux
ollama run hf.co/ggml-org/Step-3.5-Flash-GGUF:Q4_K - llama.cppMac, Windows, Linux
llama-server -hf ggml-org/Step-3.5-Flash-GGUF:Q4_K - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/ggml-org/Step-3.5-Flash-GGUF:Q4_K
On a PC Graphics Card
| Video memory | Largest size that fits | Download | Memory at 8K |
|---|---|---|---|
| 6 GB | None worth running | needs 119 GB | |
| 8 GB | None worth running | needs 119 GB | |
| 10 GB | None worth running | needs 119 GB | |
| 11 GB | None worth running | needs 119 GB | |
| 12 GB | None worth running | needs 119 GB | |
| 16 GB | None worth running | needs 119 GB | |
| 20 GB | None worth running | needs 119 GB | |
| 24 GB | None worth running | needs 119 GB | |
| 32 GB | None worth running | needs 119 GB | |
| 48 GB | None worth running | needs 119 GB |
The largest size, of those that keep three bits a weight or more, whose memory at an 8K chat is under 92% of the card's; the rest is the desktop's. Estimated from the files, the same way as for a Mac; speed on a PC is not estimated yet. A size that doesn't fit the card still runs, slower, with part of it in ordinary memory.
On every Mac
| Mac | Size | Fit | Memory | Writes |
|---|---|---|---|---|
| M6 · 16 GB | Q4_K | Too big | 126 GB | |
| M6 · 24 GB | Q4_K | Too big | 126 GB | |
| M6 · 32 GB | Q4_K | Too big | 126 GB | |
| M5 · 16 GB | Q4_K | Too big | 126 GB | |
| M5 · 24 GB | Q4_K | Too big | 126 GB | |
| M5 · 32 GB | Q4_K | Too big | 126 GB | |
| M5 Pro · 24 GB | Q4_K | Too big | 126 GB | |
| M5 Pro · 48 GB | Q4_K | Too big | 126 GB | |
| M5 Pro · 64 GB | Q4_K | Too big | 126 GB | |
| M5 Max · 36 GB | Q4_K | Too big | 126 GB | |
| M5 Max · 48 GB | Q4_K | Too big | 126 GB | |
| M5 Max · 64 GB | Q4_K | Too big | 126 GB | |
| M5 Max · 128 GB | Q4_K | Too big | 126 GB | |
| M5 Ultra · 96 GB | Q4_K | Too big | 126 GB | |
| M5 Ultra · 256 GB | Q4_K | Runs great | 126 GB | 46 |
| M5 Ultra · 512 GB | Q4_K | Runs great | 126 GB | 46 |
| M4 · 16 GB | Q4_K | Too big | 126 GB | |
| M4 · 24 GB | Q4_K | Too big | 126 GB | |
| M4 · 32 GB | Q4_K | Too big | 126 GB | |
| M4 Pro · 24 GB | Q4_K | Too big | 126 GB | |
| M4 Pro · 48 GB | Q4_K | Too big | 126 GB | |
| M4 Pro · 64 GB | Q4_K | Too big | 126 GB | |
| M4 Max · 36 GB | Q4_K | Too big | 126 GB | |
| M4 Max · 48 GB | Q4_K | Too big | 126 GB | |
| M4 Max · 64 GB | Q4_K | Too big | 126 GB | |
| M4 Max · 96 GB | Q4_K | Too big | 126 GB | |
| M4 Max · 128 GB | Q4_K | Too big | 126 GB | |
| M3 · 8 GB | Q4_K | Too big | 126 GB | |
| M3 · 16 GB | Q4_K | Too big | 126 GB | |
| M3 · 24 GB | Q4_K | Too big | 126 GB | |
| M3 Pro · 18 GB | Q4_K | Too big | 126 GB | |
| M3 Pro · 36 GB | Q4_K | Too big | 126 GB | |
| M3 Max · 36 GB | Q4_K | Too big | 126 GB | |
| M3 Max · 48 GB | Q4_K | Too big | 126 GB | |
| M3 Max · 64 GB | Q4_K | Too big | 126 GB | |
| M3 Max · 96 GB | Q4_K | Too big | 126 GB | |
| M3 Max · 128 GB | Q4_K | Too big | 126 GB | |
| M3 Ultra · 96 GB | Q4_K | Too big | 126 GB | |
| M3 Ultra · 256 GB | Q4_K | Runs great | 126 GB | 25 |
| M3 Ultra · 512 GB | Q4_K | Runs great | 126 GB | 25 |
| M2 · 8 GB | Q4_K | Too big | 126 GB | |
| M2 · 16 GB | Q4_K | Too big | 126 GB | |
| M2 · 24 GB | Q4_K | Too big | 126 GB | |
| M2 Pro · 16 GB | Q4_K | Too big | 126 GB | |
| M2 Pro · 32 GB | Q4_K | Too big | 126 GB | |
| M2 Max · 32 GB | Q4_K | Too big | 126 GB | |
| M2 Max · 64 GB | Q4_K | Too big | 126 GB | |
| M2 Max · 96 GB | Q4_K | Too big | 126 GB | |
| M2 Ultra · 64 GB | Q4_K | Too big | 126 GB | |
| M2 Ultra · 128 GB | Q4_K | Too big | 126 GB | |
| M2 Ultra · 192 GB | Q4_K | Runs great | 126 GB | 25 |
| M1 · 8 GB | Q4_K | Too big | 126 GB | |
| M1 · 16 GB | Q4_K | Too big | 126 GB | |
| M1 Pro · 16 GB | Q4_K | Too big | 126 GB | |
| M1 Pro · 32 GB | Q4_K | Too big | 126 GB | |
| M1 Max · 32 GB | Q4_K | Too big | 126 GB | |
| M1 Max · 64 GB | Q4_K | Too big | 126 GB | |
| M1 Ultra · 64 GB | Q4_K | Too big | 126 GB | |
| M1 Ultra · 128 GB | Q4_K | Too big | 126 GB |
Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.
Information
- Maker
- StepFun
- Released
- February 2026
- Parameters
- 197B, 12B active
- Longest chat
- 256K tokens
- Reads pictures
- No
- Licence
- apache-2.0
- Downloads this month
- 13,020
- Likes
- 15
- Sizes
- 1, 119 to 119 GB
- Kind
- Maker's release
Badge for the model card
Shows on Step 3.5 Flash's Hugging Face card which Macs run it well, linked to this page. It updates every morning.
Files, downloads and likes read from Hugging Face on November 5, 2026 and refreshed every morning. Memory and speed are estimates until someone running Outmake measures this model on that Mac.