All tasks / Models
Step-Audio-EditX
Open-weights 3B model from StepFun for expressive text-to-speech and iterative editing of emotion and style.
- StepFun
- Open weights
- Released 29 Oct 2025
Where it ranks
- CText to speechModels · Local · 1,100 on Artificial Analysis Speech Arena: Text to Speech (provider voices)
Facts
- Vendor
- StepFun
- Family
- Step-Audio-EditX
- Released
- 29 Oct 2025
- Weights
- Open: download and run it yourself
- License
- Apache-2.0
- Parameters
- 3.53B
- GPU memory to run locally
- ~8 GB (Derived: about 3.53B parameters is roughly 7.1 GB of bf16 weights plus runtime overhead; not an official requirement.)
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Benchmark results
| Benchmark | Result | Entry on the leaderboard | As of | Source |
|---|---|---|---|---|
| Artificial Analysis Speech Arena: Text to Speech (provider voices) | 1,100 | Step Audio EditX (Mar 2026)Provider Voice Arena, all categories and accents; rank 39 of 96, 2249 appearances; raw value from the page's embedded data; no last-updated date shown, so as_of = retrieval date. | 19 Oct 2026 | artificialanalysis.ai |
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.