# StepFun 3.5 Flash

An open-weight reasoning model that pairs strong math and coding scores with a small active-parameter footprint. StepFun 3.5 Flash targets agents that need to reason deeply, call tools, and stay cheap to run.

- **Model ID:** `stepfun/step-3.5-flash`
- **Type:** chat
- **Providers:** deepinfra
- **Context window:** 262,114
- **Maximum output tokens:** 262,114
- **Pricing:** $0.09/1M input tokens, $0.3/1M output tokens
- **Canonical page:** https://vercel.com/ai-gateway/models/step-3.5-flash

## Supported parameters

Detailed capability metadata has not been reported for this model.

## Example

```ts
import { streamText } from 'ai'

const result = streamText({
  model: 'stepfun/step-3.5-flash',
  prompt: 'Why is the sky blue?'
})
```

## About

StepFun 3.5 Flash is StepFun's open-source reasoning model, released under the Apache 2.0 license with weights published on GitHub and Hugging Face. The architecture is a sparse mixture of experts with 196B total parameters and about 11B active per token. Each token routes through eight of 288 experts plus one shared expert, so inference cost tracks the active subset rather than the full parameter count.

Two design choices keep long-context work affordable. A 3:1 ratio of sliding-window to full attention layers supports the 262.1K tokens context window without quadratic cost across every layer. Multi-token prediction lets StepFun 3.5 Flash draft several tokens per forward pass, which accelerates generation.

StepFun 3.5 Flash posts frontier-level reasoning scores for its size: 97.3 on AIME 2025, 74.4% on SWE-bench Verified, 86.4% on LiveCodeBench-v6, and 88.2% on Tau2-Bench. Deep reasoning, tool calling, and agentic control loops are first-class capabilities rather than add-ons.

Through AI Gateway, you call StepFun 3.5 Flash with a single API key and get provider routing, automatic failover, and built-in observability. Integrate via the AI SDK, the Chat Completions API, the Responses API, the Messages API, or other supported API formats. No StepFun account is required.

## What to consider

StepFun 3.5 Flash is a reasoning model, so responses include thinking tokens before the final answer. Budget output tokens accordingly, and tune reasoning depth where your harness allows it.

StepFun 3.5 Flash is text-only. Route requests that carry images or video to a multimodal model instead. Benchmark figures above come from StepFun's published evaluations, so validate on your own workload before committing production traffic. For current throughput and latency numbers, see live metrics on this page.

## When to use

### Best For

- Agentic pipelines that chain deep reasoning with multi-step tool calls
- Math-heavy and competition-style reasoning, backed by a 97.3 AIME 2025 score
- Coding agents that want a 74.4% SWE-bench Verified model with a small active footprint
- Long-context repository or document analysis across the full 262.1K tokens window
- High-volume production traffic where sparse activation keeps inference costs down

### Consider Alternatives When

- Your inputs include images or video, since StepFun 3.5 Flash is text-only
- You need the highest scores on the hardest coding and research benchmarks
- Simple single-turn chat doesn't benefit from reasoning-token overhead

## Best for

- **Agentic Tool Loops:** Pipelines that chain deep reasoning with multi-step tool calls
- **Math and Reasoning Workloads:** Competition-style problems backed by a 97.3 AIME 2025 score
- **Cost-Efficient Coding Agents:** Software engineering tasks served by a 74.4% SWE-bench Verified model with only 11B active parameters
- **Long-Context Analysis:** Repository or document review that uses the full 262.1K tokens window
- **High-Volume Production Traffic:** Workloads where sparse MoE activation keeps per-request costs low

## Consider alternatives

- **Multimodal Inputs:** StepFun 3.5 Flash is text-only, so route image or video requests to a vision-capable model
- **Maximum Benchmark Headroom:** Larger frontier models still lead on the hardest coding and research suites
- **Simple Single-Turn Chat:** Reasoning tokens add output overhead that lightweight conversation doesn't need

## Frequently asked questions

### What architecture does StepFun 3.5 Flash use?

StepFun 3.5 Flash uses a sparse mixture-of-experts design with 196B total parameters and about 11B active per token. Each token routes through eight of 288 experts plus one shared expert, and a 3:1 sliding-window attention ratio keeps long-context inference efficient. See https://deepinfra.com/stepfun-ai/Step-3.5-Flash for details.

### Is StepFun 3.5 Flash open source?

Yes. StepFun released StepFun 3.5 Flash under the Apache 2.0 license, with weights available on GitHub and Hugging Face. Through AI Gateway you use hosted inference, so licensing matters only if you also self-host.

### What is the context window for StepFun 3.5 Flash?

StepFun 3.5 Flash supports a context window of 262.1K tokens and a max output of 262.1K tokens per request.

### How well does StepFun 3.5 Flash perform on benchmarks?

StepFun 3.5 Flash scores 97.3 on AIME 2025, 74.4% on SWE-bench Verified, 86.4% on LiveCodeBench-v6, and 88.2% on Tau2-Bench in StepFun's published evaluations. Validate on your own workload before shipping.

### How do I call StepFun 3.5 Flash through AI Gateway?

Use the model identifier `stepfun/step-3.5-flash` with the AI SDK, the Chat Completions API, the Responses API, the Messages API, or another supported API format. You authenticate with an AI Gateway API key, and no StepFun account is needed.

### Does AI Gateway support Zero Data Retention for StepFun 3.5 Flash?

Yes, Zero Data Retention is available for this model. Zero Data Retention is offered on a per-provider basis. See https://vercel.com/docs/ai-gateway/capabilities/zdr for details.

### Can StepFun 3.5 Flash handle image or video inputs?

No. StepFun 3.5 Flash is text-only. For visual inputs, pick a multimodal model from the AI Gateway catalog, such as a later Step-series flash release with native image understanding.

## Links

- [Model page](https://vercel.com/ai-gateway/models/step-3.5-flash)
- [AI Gateway documentation](https://vercel.com/docs/ai-gateway)
- [Provider model documentation](https://deepinfra.com/stepfun-ai/Step-3.5-Flash)