# Step 3.5 Flash

> Step 3.5 Flash is a sparse Mixture-of-Experts model by StepFun with 196.81B total parameters (196B backbone + 0.81B MTP head) and ~11B active per token. Built on a 45-layer transformer with 288 routed experts (Top-8 selection), 3:1 SWA attention ratio, and 256K context. Achieves 100–300 tok/s throughput (peaking at 350 tok/s for coding) for frontier reasoning and agentic tasks.

**ID**: `stepfun-ai/step-3.5-flash`  
**Creator**: stepfun-ai  
**Category**: text  
**Context**: 256K tokens  
**Pricing**: $0.10 / $0.30 per 1M · billed by your provider (BYOK)  
**Released**: 2026-06-01  
**Web page**: https://anyrouter.dev/model/stepfun-ai/step-3.5-flash

**Input modalities**: text  
**Output modalities**: text  

**Tokenizer**: StepFun

**Capabilities**: chat, streaming, coding, function-calling

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/chat/completions`

```bash
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "stepfun-ai/step-3.5-flash",
  "messages": [
    {
      "role": "user",
      "content": "Say hi in 3 words."
    }
  ]
}'
```

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| nousresearch-byok | $0.10/M (list) | $0.30/M (list) |
| stepfun-byok | $0.10/M (list) | $0.30/M (list) |
| stepplan-byok | $0.10/M (list) | $0.30/M (list) |
| commandcode-byok | $0.10/M (list) | $0.30/M (list) |

BYOK routes are billed by your own provider key — AnyRouter charges $0. Rates marked (list) are the provider's published list price.

## Benchmarks

- **GPQA Diamond**: 0.8
- **HLE**: 0.2
- **τ²-Bench Telecom**: 0.9
- **SciCode**: 0.4

## Source

- Upstream docs: https://build.nvidia.com/stepfun-ai/step-3.5-flash