Step 3.7 Flash

Step 3.7 Flash is a newer StepFun flash model for faster agents, coding, and multimodal prompts. It is part of the step line, following Step 3.5 Flash 2603. It has a context window of 256000 tokens and a max output of 256000 tokens. Weights are open. It supports reasoning mode, tool calling, and accepts image, text, and video, returning text. Knowledge cutoff is 2026-03-01. Price is from $0.185 per million input tokens and $1.11 per million output tokens, with cached input from $0.03 per million tokens. It is served by 19 hosts.

Against Step 3.5 Flash 2603: input $0.10 to $0.185, output $0.30 to $1.11, cached input $0.02 to $0.03.

Range across the hosts that serve it: up to $0.22 per 1M input tokens.

• ·Fast agents and coding tasks
• ·Multimodal input: image, text, video
• ·256k context and output tokens
• ·Open weights with reasoning and tool calling
• ·Cached input at $0.03 per million tokens

• Reach for it for building agent workflows
• Reach for it for coding assistance
• Reach for it for multimodal prompts with video
• Reach for it for long context tasks

19 hosts serve this model at the price above · the maker's documentation

• 8Artificial Analysis 评测 Step 5 Preview： Intelligence Index 得 44 分，成本约为同级模型 1/2.82026-09-22

1 signal of the archive mention this model, each scored against the published bar. This list is rebuilt from the archive at every build, so it cannot go stale.

The registry keeps the six most recent versions of a line. Older ones stay in the file and can be surfaced without a rebuild.

What is the price per token?

Input from $0.185 per million, output $1.11 per million, cached input $0.03 per million.

Does it support tool calling?

Yes, it supports tool calling and reasoning mode.

What input types does it accept?

It accepts image, text, and video, and returns text.