StepFun
Chat with Step 3.7 Flash online
StepFun's efficiency model — Apache 2.0, fast enough to call in a loop and capable enough to trust.
262.144K tokens
컨텍스트
131,072 tokens
최대 출력
빠름
속도
이럴 때 활용하세요
High-volume agents
Loops where latency and price compound quickly.
Everyday coding
Fast answers on real code without the frontier price.
Search workflows
Reading and filtering a lot of material quickly.
Structured output
Clean JSON when the answer feeds a pipeline.
Prompts to steal
Tuned to this model — click any line to copy.
Step 3.7 Flash 소개
Step 3.7 Flash is StepFun's efficiency model, released on 28 May 2026 under an Apache 2.0 licence. It is a sparse mixture-of-experts design of about 198 billion parameters that activates only around 11 billion per token, which is the whole point: it answers quickly and costs little, while staying capable enough for real work.
StepFun built it for coding agents and search-style workflows — the kind of job where a model is called many times in a loop and both latency and price compound. Throughput runs to several hundred tokens a second on a good rail, and the model exposes selectable reasoning levels so a caller can trade depth against speed per request rather than being locked into one setting. It takes tools and tool_choice for function calling and supports a JSON response format when the output feeds a pipeline.
The value claim is the interesting part. On SWE-Bench Verified with its advisor mode enabled, StepFun reports Step 3.7 Flash reaching around 97% of a frontier model's coding score at roughly a ninth of the per-task cost. Treat any single benchmark carefully, but the shape of the claim matches what the architecture is for.
The context window is 262,144 tokens, so a substantial codebase or a long document set fits in one conversation. On NinjaChat it is included in every plan — a good default when you want a capable model to answer fast and often.
Step 3.7 Flash 사용 방법
1분 안에 첫 결과물까지.
01
Create a NinjaChat account and choose a plan
02
Open chat and pick Step 3.7 Flash from the model list
03
Use it as your default for fast, frequent questions
04
Escalate to a frontier model when a task gets hard
더 나은 결과를 위한 팁
- Great as the first pass before a bigger model
- Ask for JSON when the answer feeds something else
- Its context is large; paste the whole file
- Pair it with a frontier model rather than replacing one
Step 3.7 Flash vs 다른 대안들
{model}이 앞서는 부분과 그렇지 않은 부분, 솔직하게 비교합니다.
vs Qwen 3.8 Flash
모델 보기Larger context and selectable reasoning levels
Qwen 3.8 Flash is cheaper still
vs GLM 4.7
모델 보기Cheaper per token at a similar everyday coding job
GLM 4.7 has the wider GLM ecosystem behind it
vs Kimi K2.7 Code
모델 보기Much faster and cheaper for routine work
Kimi K2.7 Code is far stronger on long agent runs
