StepFun · Text
StepFun models: Versions, specs, GPU requirements & hosting
StepFun separates Flash, general, vision and audio models so modality-specific releases do not distort generation chronology.
Overview and history
StepFun is an official StepFun model family represented here through independently verified release, checkpoint, license, runtime and hardware evidence. StepFun separates Flash, general, vision and audio models so modality-specific releases do not distort generation chronology. Current public releases are grouped by the Flash, General, Vision, Audio tracks so sibling sizes and specialist branches are not presented as false chronological generations.
FusionPointAI currently tracks 12 certified public StepFun releases. The catalogue preserves release dates and evaluates Current, Recent, Previous and Legacy status within each model track rather than by database insertion order.
Fast inference models
Vision and video understanding
Audio reasoning
Architecture and model range
The verified public range spans 8B to 198B total parameters and context windows up to 262,144 tokens. Each release states whether it is dense or mixture-of-experts and records active parameters separately when the official source publishes them.
Public releases12
Model tracks3
Current generationStep 3.7 Flash
Last verified2026-09-04T00:00:00.000000Z
Generations, tracks and variants
Flash
step-3-5
Recent generation 3 sibling variantsstep-3-7
Current generation 5 sibling variantsStep-3.7-Flash-NVFP4
Not published by manufacturer parameters · N/A / Not directly comparable context
Base
Step-3.7-Flash-GGUF
Not published by manufacturer parameters · N/A / Not directly comparable context
Quantization
Step-3.7-Flash
Not published by manufacturer parameters · N/A / Not directly comparable context
Base
Step-3.7-Flash-FP8
Not published by manufacturer parameters · N/A / Not directly comparable context
Quantization
Step 3.7 Flash
198.0B parameters · 262,144 context
Fast · MoE
MoE Transformer
Vision
Step3-VL
Current generationAudio
step-audio-2
Recent generation 2 sibling variantsStep-Audio R1.1
Current generationDeployment ecosystem
Choose a specific StepFun release and certified checkpoint before sizing hardware. FusionPointAI rejects unknown runtime configurations, calculates resident memory from total model weights, then distinguishes globally compatible GPUs from currently rentable inventory.
Validate the exact checkpoint license, runtime version, precision kernel, context length, concurrency and multi-GPU topology. Active MoE parameters affect compute, but resident model memory continues to use total stored weights.
Common use cases
Low-latency assistants
Multimodal analysis
Speech applications
License overview
Licenses are certified per release, not inferred from the provider. Review the release page before commercial deployment because terms can differ between generations and variants.
Commercial-use status is a source-backed catalogue classification, not legal advice. Follow the linked official license and attribution requirements.
Release timeline
stepfun/stepfun/step-37-flash-fp8 @ c9a8e60ebb37be7e193690fcefd13b1de5f7f8ba536a387c67a7cf34e7883a0b
stepfun/stepfun/step-37-flash-nvfp4 @ a764b426a44f9833723b8953ea0f756a6646d512a4087b352e60c3eb6579917b
stepfun/stepfun/step-audio-r1-1 @ 11a5ae11040a8f5b7cbc16a1bb4bf1be3b9871a0ea308b8e57a146ecd767ed1e
stepfun/stepfun/stepfun-step-audio-2-mini @ b12a2b1f7790db6c5705fe90bf0b39b8ce143905bbf9ba090c96cd0f312df613
stepfun/stepfun/step-35-flash @ 4af706d08154a632d5497e8fbc66558132b6245ada0ad125a6f4a312443b30ea
stepfun/stepfun/step-37-flash @ 0175fba75c7b40c24436c590cd70792a204451388ddc80e092c4d9e5f6a77de9
stepfun/stepfun/step-37-flash-gguf @ 08ae99dd0e4ae6b110301ef32d9737926f02c210f3db054caf5ad091b065f34d
stepfun/stepfun/step-3-7-flash @ e0affc198b51388bfb7e8fa14a6f692b11fa1b28099334a262d2ca8a3b4552ac
stepfun/stepfun/step-3-5-flash @ 861dcb086e44521b2c16b56c08fa7c05a3c19e06151bf96b75fba2ac6e1637c0
stepfun/stepfun/step-audio-2-mini @ 78fc4999e3d019c52880daa8c85099216f3e3ef53583974d3642f2c76bf2b106
stepfun/stepfun/step3-vl-10b @ 6834d893eb35cd0c0af8e24c76c9de39b786c778637069212687e46a5b135c0a
stepfun-ai/Step-3.5-Flash @ 2026-07-10T12:06:23Z
stepfun-ai/Step-3.7-Flash @ 2026-06-03T12:37:55Z
stepfun-ai/Step-Audio-2-mini @ 2026-02-14T12:15:51Z
Frequently asked questions
Which StepFun release is current?
Current releases are Step 3.7 Flash, Step3-VL 10B, Step-Audio R1.1, Step-3.7-Flash, evaluated within their respective tracks.
Can StepFun be self-hosted?
Only releases with a verified official checkpoint and supported runtime configuration are presented as deployable.
Which GPU should run StepFun?
GPU selection depends on the exact variant, checkpoint precision, context length and concurrency; use the release hardware recommendations rather than the family name alone.
Official references
- huggingface.co
- huggingface.co
- huggingface.co
- huggingface.co
- huggingface.co
- modelscope.cn
- modelscope.cn
- modelscope.cn
- huggingface.co
- huggingface.co
- huggingface.co
- huggingface.co
Last verified: 2026-09-04