StepFunStep 3.7 Flash
Step 3.7 Flash is a 198B-parameter sparse Mixture-of-Experts vision-language model built by StepFun, combining a large language backbone with a vision encoder for native image understanding. It is engineered for high-frequency production workloads, activating approximately 11B parameters per token and supporting a 262,144 token context window.
Input
Output
Context
262K
Max Output
-
Parameters
201.4B
Technical Specifications
Input Modalities
Output Modalities
Capabilities
Loading capabilities…
Benchmarks
Estimated GPU requirements for Step 3.7 Flash
Estimates based on INT8 quantization at up to 32K context. A count above one assumes tensor parallelism across the cards. Actual requirements vary by framework and configuration.
More models by StepFun
The creator's other models in the catalog, with their context, size and license where published.
Compare it, or keep browsing
Put this model beside its alternatives on the same evidence, or go back to the full catalog.
Frequently asked questions about Step 3.7 Flash
Answered from the entry's own fields: context, license, modalities, evidence, serving and the memory to self-host.
Step 3.7 Flash advertises a context window of 262,144 tokens. The figure is the creator's published maximum; a given provider may serve less, and the gateway routes on what each provider actually serves.
Yes. Step 3.7 Flash is an open-weight model released under the Apache 2.0 license, so the weights can be downloaded and self-hosted within that license's terms.
Step 3.7 Flash accepts Text and Image and produces Text. The capabilities card on this page lists which API features each deployment honours, such as function calling and structured output, with the source each was checked against.
Step 3.7 Flash has published results from Artificial Analysis, shown by suite in the benchmarks card above exactly as the publisher reported them. Scores are not combined across suites, and a suite that has not measured Step 3.7 Flash is shown as not available rather than estimated.
Not on the managed pool today; Step 3.7 Flash is listed for reference and comparison. Connect your own provider key or endpoint that serves it and the gateway runs it on your account, with routing decisions recorded the same way.
About 198.4 GB at INT8 for the weights and a default context, from the catalog's 201.4B parameter count; FP16 needs roughly twice that, and long contexts or many concurrent requests add KV cache on top. The GPU section on this page lists cards that hold it, and the capacity planner sizes it for your context length and traffic.
Fields collected from public registries, provider APIs and benchmark publishers, each tagged with its source.
Last updated: Nov 4, 2026
Put intelligence in the middle.
One gateway in front of every model, with your policies applied and every decision on record. Start on the free plan with 5,000 requests a month, no card required.