The page lists a 1 million-token context window and maximum input, plus a maximum output of 64,000 tokens.
The model accepts text, image, and video inputs and returns text. The documentation lists streaming, tool calling, prompt caching, JSON Mode, JSON Schema, and configurable reasoning effort at low, medium, or high. It supports up to 60 images in a request. Video can be supplied by URL, Base64, or a Files API reference; individual MP4 URLs must be under 128 MB, and videos under five minutes are recommended.
StepFun points developers to Chat Completions API and Claude Code integration documentation, and says tools such as search and code execution are supplied by the integrating application rather than by the model itself. The page directs readers elsewhere for current pricing, cache billing, and account rate limits. It does not provide prices, benchmark results, regional availability terms, or independent evidence of performance.