Step 3.5 Flash

Tool callingReasoning

Hugging Face Website

StepFun flash lane for quick multimodal reasoning and coding assistance

Context: 262K
Released: 8mo ago
Parameters: —
Input / 1M tokens: $0.10
Output / 1M tokens: $0.30

Sources:models.devOpenRouter

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

No benchmark results have been matched to this model yet.

Specifications

Developer: Stepfun
Model family: Step
Type: Language model
Predecessor: —
Successor: —
Released: January 29, 2026
Model ID: stepfun/step-3.5-flash
Openness: Open weights
License: —
Parameters: —
Input modalities: text
Output modalities: text
Modality: text->text
Tokenizer: Other
Max output tokens: 66K
Knowledge cutoff: 2025-01
Hugging Face: stepfun-ai/Step-3.5-Flash
Source freshness: Metadata synced Oct 8, 2026, 10:17 PM UTC

Additional source evidence

Source model: stepfun/step-3.5-flash
Last updated: 2026-02-13
Knowledge cutoff: 2025-01
Model context: 256K
Maximum input: 256K
Model output limit: 256K
Open weights: Yes
File attachments: No
Reasoning: Yes
Tool calling: Yes
Temperature control: Yes

Downloadable weights

Hugging Face

Published benchmarks

Benchmark	Score	Evaluation details	Evidence
Artificial Analysis Coding Index	31.6 · index	2026-06-02	Source
SciCode	40.4 · percent correct	2026-06-02	Source
Terminal-Bench Hard	27.3 · success rate	2026-06-02	Source
SWE-Bench Verified	74.4 · resolved	—	Source

Complete source record (JSON)

{
  "id": "stepfun/step-3.5-flash",
  "name": "Step 3.5 Flash",
  "description": "StepFun flash lane for quick multimodal reasoning and coding assistance",
  "last_updated": "2026-02-13",
  "knowledge": "2025-01",
  "attachment": false,
  "reasoning": true,
  "tool_call": true,
  "temperature": true,
  "modalities": {
    "input": [
      "text"
    ],
    "output": [
      "text"
    ]
  },
  "limit": {
    "context": 256000,
    "input": 256000,
    "output": 256000
  },
  "benchmarks": [
    {
      "name": "Artificial Analysis Coding Index",
      "score": 31.6,
      "source": "https://openrouter.ai/stepfun/step-3.5-flash/benchmarks",
      "metric": "index",
      "date": "2026-06-02"
    },
    {
      "name": "SciCode",
      "score": 40.4,
      "source": "https://openrouter.ai/stepfun/step-3.5-flash/benchmarks",
      "metric": "percent correct",
      "date": "2026-06-02"
    },
    {
      "name": "Terminal-Bench Hard",
      "score": 27.3,
      "source": "https://openrouter.ai/stepfun/step-3.5-flash/benchmarks",
      "metric": "success rate",
      "date": "2026-06-02"
    },
    {
      "name": "SWE-Bench Verified",
      "score": 74.4,
      "source": "https://arxiv.org/abs/2602.10604",
      "metric": "resolved"
    }
  ],
  "release_date": "2026-01-29",
  "open_weights": true,
  "weights": [
    {
      "label": "Hugging Face",
      "url": "https://huggingface.co/stepfun-ai/Step-3.5-Flash"
    }
  ]
}

Infrastructure

Provider availability

All providers

Provider	Endpoint	Context	Input /M tokens	Output /M tokens	Cached input /M tokens	Cache write /M tokens	Uptime 1d	Data handling
SiliconFlow	siliconflow/fp8	262K	$0.10	$0.30	—	—	99.39%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15

Provider data:OpenRouter

Deployment graph

Can I run this?

Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.

Browse hardware