Instructions to use stepfun-ai/Step-3.5-Flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use stepfun-ai/Step-3.5-Flash with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="stepfun-ai/Step-3.5-Flash", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("stepfun-ai/Step-3.5-Flash", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("stepfun-ai/Step-3.5-Flash", trust_remote_code=True, device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use stepfun-ai/Step-3.5-Flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "stepfun-ai/Step-3.5-Flash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "stepfun-ai/Step-3.5-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/stepfun-ai/Step-3.5-Flash

SGLang

How to use stepfun-ai/Step-3.5-Flash with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "stepfun-ai/Step-3.5-Flash" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "stepfun-ai/Step-3.5-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "stepfun-ai/Step-3.5-Flash" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "stepfun-ai/Step-3.5-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use stepfun-ai/Step-3.5-Flash with Docker Model Runner:
```
docker model run hf.co/stepfun-ai/Step-3.5-Flash
```

SaylorTwift HF Staff commited on 7 days ago

Commit

688ca5a

verified ·

1 Parent(s): ab446a3

Add IFBench evaluation result

Browse files

# Add IFBench evaluation result

## Summary

This PR adds the IFBench score reported for `stepfun-ai/Step-3.5-Flash` to `.eval_results/`, following the [Hugging Face Hub evaluation-results specification](https://huggingface.co/docs/hub/eval-results). IFBench ([allenai/IFBench_test](https://huggingface.co/datasets/allenai/IFBench_test)) is now an official Hub benchmark, with prompt-level loose accuracy as its headline metric.

## Benchmarks Added

- [IFBench](https://huggingface.co/datasets/allenai/IFBench_test?eval_result=stepfun-ai/Step-3.5-Flash&leaderboard_task_id=prompt_level_loose): 67.4

## Source

- Step 3.5 Flash technical report: https://arxiv.org/abs/2602.10604 (row: "IFBench (Vanilla)")

## Files Added

- `.eval_results/ifbench.yaml`

## Verification

This score is self-reported in the source above; no verified token is provided because it was not run via HF Jobs with inspect-ai. The Hub task is prompt-level loose accuracy (0-100). If your number was computed with a different IFBench variant (strict, or instruction-level), let us know and we will fix or close this PR.

Files changed (1) hide show

.eval_results/ifbench.yaml +8 -0

.eval_results/ifbench.yaml ADDED Viewed

	@@ -0,0 +1,8 @@

+- dataset:
+    id: allenai/IFBench_test
+    task_id: prompt_level_loose
+  value: 67.4
+  date: "2026-03-17"
+  source:
+    url: https://arxiv.org/abs/2602.10604
+    name: "Step 3.5 Flash technical report"