Stepfun

SStep 3.5 FlashOPEN

Step 3.5 Flash (Stepfun): live API pricing, context window and the best open-source alternatives, tracked daily by olud.ai.

Context window
262K
tokens
Input price
$0.1
per M tokens
Output price
$0.3
per M tokens
Provider
Stepfun

Prices update automatically — checked hourly against provider list prices.

See model comparisons → Compare all model prices

Benchmarks & performance

Independent benchmark scores for Step 3.5 Flash, measured by Artificial Analysis. Higher is better.

28×
More intelligence per dollar than Claude Opus 5 (Fast).
For the same budget, Step 3.5 Flash delivers 28 times more capability. Open weights also mean you can self-host it and pay nothing per token.
Intelligence index26
GPQA82.6%
Humanity's Last Exam22.6%
Long Context Reasoning54.3%
SciCode38.5%
IFBench66.5%
τ²-Bench87.4%
Terminal-Bench Hard32.6%
⚡ Speed289.4 tokens/sec
⏱ Latency0.75s to first token
💰 Blended price$0.15 / 1M tokens
📈 Value173.3 intelligence points per $
Benchmark data by Artificial Analysis

About this model

Step 3.5 Flash is an open-weight AI model by Stepfun. You can download and self-host it for free; the prices below are hosted-API list prices, tracked hourly, for when you prefer convenience over self-hosting.

Frequently asked questions

What is Step 3.5 Flash?

Step 3.5 Flash is an AI language model from Stepfun. It is open-weight: you can download it and run it on your own hardware, for free. It scores 26 on the Artificial Analysis intelligence index.

Is Step 3.5 Flash free?

The weights are free and open — you can self-host Step 3.5 Flash and pay nothing per token. If you prefer a hosted API, list prices are $0.1 per million input tokens and $0.3 per million output tokens.

What is Step 3.5 Flash good at?

Independent benchmarks from Artificial Analysis give it GPQA 82.6%, Humanity's Last Exam 22.6%, Long Context Reasoning 54.3%, SciCode 38.5%, IFBench 66.5%, τ²-Bench 87.4%, Terminal-Bench Hard 32.6%.

How fast is Step 3.5 Flash?

It generates about 289.4 tokens per second, with a median 0.75s delay before the first token. Measured independently by Artificial Analysis.

Can I self-host Step 3.5 Flash?

Yes. Step 3.5 Flash has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.

Related models

Step 3.7 FlashStepfun