StepFun: Step 3.7 Flash

Step 3.7 Flash is a StepFun model that accepts text, image, and video input. It offers a 262,144-token context window with up to 230,400 completion tokens, supports tool calling, and includes reasoning support. Structured output support is not listed, so teams that rely on strict schemas should verify it before committing. Pricing is $0.20 per million input tokens and $1.15 per million output tokens. Because input costs far less than output, the model fits read-heavy workloads such as analyzing long documents, images, or video, where most tokens go in rather than out. The large completion ceiling allows long outputs, though those raise costs quickly. There is no independent benchmark coverage yet, so quality remains unproven. Shortlist it for multimodal, tool-using pipelines where input cost matters, and run your own evaluations against established alternatives before depending on it.