Kyma APIKyma API
StepFun

StepFun

Step 3.7 Flash

Step 3.7 Flash is a fast, low-cost model optimized for high-throughput chat and bulk processing. Reach for it when you need to handle text, image, or video inputs at scale with tool calling and reasoning enabled.

Modalities

Text+Image → Text

Below release

fp8

Weights

Published

stepfun-ai/Step-3.7-Flash

Where it sits in the catalogue Kyma measures

Every number against every text model Kyma prices per token — a stated rule, not a chosen line-up.

Price · input#17/71

$0.27

output $1.552 · cached $0.054

median $1.013 · best $0.0494

Price · output#21/70

$1.552

per 1M tokens generated

median $3.531 · best $0.218

Availability · 30d#1/54

100%

102 observations since 2026-06-10

median 99.9% · best 100.0%

Throughput#20/65

53 tok/s

probe median, one fixed prompt

median 35.7 tok/s · best 365.1 tok/s

Response time#36/65

3.04 s

probe median, to a complete answer

median 2.96 s · best 0.89 s

Context

256K

max output 8,192

published by StepFun, not measured here

Tick above each rail is this model, below it the other 70. Dashed rule is the field median, solid is its best. Better is left; the axis stops at the 90th percentile, so a few models sit past its right edge.

Usage

How much this model is actually called here.

Rank

#47

of 101 active models

Tokens served

1.3M

all-time

Platform share

0.0%

of all tokens

Tokens · last 14 daysSep 26 → Oct 9

Two clocks, and why they disagree

Kyma measures this model twice. Both are real; they answer different questions.

Probe · every 6h · 30 days

3.04s to answer

One fixed prompt, on a schedule, to every model. Comparable, because the model is the only thing that changes.

Observations
102
Answered by a substitute
0

Real traffic · last 7 days

2.83s to answer

Your requests, at the lengths clients actually send. Not comparable between models, but it is what running this one feels like.

Requests
44
Completed
100%
p95
5.1 s

Caching, as realised · last 7 days

0.0% of input cached

The share of input that actually hit cache, so the effective rate below is what was charged, not a best case.

Cached input
0
Fresh input
619
List input
$0.27 /1M
Effective input
$0.27 /1M

Production prompts here are shorter than the probe's. Use the probe figure to choose between models, the traffic figure to budget for your own.

Pricing

Pay per token. Cached input bills at this model’s own cached rate, listed below.

$0.27 /1M input$1.552 /1M output
Hobby10 req/day · 2K in / 500 out
~$0.39/mo
Production1,000 req/day · 2K in / 500 out
~$39.48/mo
Scale20,000 req/day · 2K in / 500 out
~$790/mo
+ Estimate your workload
1,000
2,000
500
30%

Estimated monthly cost

$35.59

$1.19 / day on Step 3.7 Flash

Same workload on:

Qwen 3.8 27B$59.77+68%
Llama 4 Maverick$32.40-9%

Estimates use list pricing with cached input at this model's own cached rate. Actual bills depend on real token counts, and every response includes its exact cost.

When to use Step 3.7 Flash

Updated 2026-07-31

Where this model earns its cost — and where it doesn't.

Created by stepfun, it supports a 256,000-token context window and accepts text, image, and video inputs while returning text. It includes native support for tool calling, reasoning, and structured outputs.

On Kyma, requests are routed through an OpenAI-compatible endpoint with automatic failover if a serving path degrades. Prompt prefixes are cached at this model's cached input rate.

The model caps output at 8,192 tokens, making it suitable for conversational and extraction workflows rather than long-form generation.

Bulk Chat Processing

Route high-volume conversational workloads with low latency and minimal overhead.

Multimodal Input Analysis

Extract information from text, images, and video streams in a single request.

Tool-Driven Workflows

Execute function calls and structured data extraction for automated pipelines.

Long-Context Summarization

Process documents up to 256K tokens while maintaining fast response times.

Not ideal for: It is not suited for tasks requiring outputs longer than 8,192 tokens or heavy creative writing where maximum reasoning depth is prioritized over speed and cost.

How it compares

Against the peers people actually weigh it against.

SpecStep 3.7 FlashQwen 3.8 27BLlama 4 Maverick
Input /1M$0.27$0.3576$0.27
Output /1M$1.552$2.554$1.08
Context256K1M1M
ToolsYesYesYes
ReasoningYesYesNo
Throughput53 tok/s36 tok/s34 tok/s

Quick start

Up and running in under two minutes.

  1. 1

    Create an API key

    Sign up and grab a key from the dashboard — $0.50 free credit on the free tier, which covers Step 3.7 Flash. No card required.

    Get API key →
  2. 2

    Make your first request

    Drop in your key and send a chat completion — fully OpenAI-compatible.

    curl https://kymaapi.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "step-3.7-flash",
    "messages": [
      {"role": "user", "content": "Explain prompt caching in one paragraph."}
    ]
  }'
  3. 3

    Stream responses

    Add "stream": true to receive tokens as they arrive.

    curl https://kymaapi.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "step-3.7-flash",
    "stream": true,
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

FAQ

Common questions about this model.

What is the context window of Step 3.7 Flash?

Step 3.7 Flash has a 256K-token context window — roughly 376 pages of text in a single request.

How much does the Step 3.7 Flash API cost?

$0.27 per 1M input tokens and $1.552 per 1M output tokens, with cached input at $0.054/1M — 80% off the input rate on repeated prompt prefixes. No subscription; you pay only for what you use.

Does Step 3.7 Flash support function calling?

Yes — Step 3.7 Flash supports tool/function calling and structured outputs (JSON mode), so it works with agent frameworks out of the box.

Are the weights for Step 3.7 Flash publicly available?

Yes. StepFun publishes Step 3.7 Flash's weights as stepfun-ai/Step-3.7-Flash, so you can download and run the model yourself (https://huggingface.co/stepfun-ai/Step-3.7-Flash, read 2026-08-14). Kyma serves it because it is convenient and has failover behind it, not because it is the only way to reach it.

Is Step 3.7 Flash ever served below the precision its creator released it at?

Sometimes. Step 3.7 Flash was released by StepFun at bf16 + fp32 (https://huggingface.co/stepfun-ai/Step-3.7-Flash, read 2026-08-14), and at least one route serving it here reports fp8 — narrower than that. Weights compressed below the release usually answer close to it, but it is not the identical artefact, and nothing in the response tells you which one answered. Kyma checked the routes on 2026-09-21, so it is your call rather than a silent one.

How do I use Step 3.7 Flash?

Kyma is OpenAI-compatible: point your SDK's base URL at https://kymaapi.com/v1, use your Kyma API key, and set the model to step-3.7-flash. Signing up is free and includes $0.50 of credit on the free tier, which covers this model — no card required.

Does this model support image and video inputs?

Yes, it accepts text, image, and video inputs and returns text outputs.

How does prompt caching work with this model on Kyma?

Repeated prompt prefixes are billed at this model's cached input rate, and Kyma automatically reroutes requests if a serving path degrades.

What is the maximum output length?

The model caps generation at 8,192 tokens per request.

Start with $0.50 free credit on the free tier — no card required.Create account →