🚶 StepFun

Proprietary Step models — fast inference across chat, vision, and STT/TTS

🚶

StepFun

Speed: ⚡ 93 tok/s avg
Models: 6
Price: $9/mo Step Plan
Status: ✅ Online
Avg Speed
93 tok/s
Models
6
Price
$9/mo
Best For
Step models, STT/TTS

💰 Plan & Pricing

ℹ️
Step Plan — $9/mo
Access to all Step models including chat, vision, reasoning, STT, and TTS. Own models only (step-3.x series).

🔑 API Key

4d2tfWopcoBE9gHPP5boAGiYCrFB7iFiFCt1EluOGoWLVlQVxePCvNh6MwN5I7MKE

🌐 Endpoint

https://api.stepfun.ai/step_plan/v1/chat/completions

📦 Models (6 LLMs + 2 Audio)

ModelSpeedCategoryNotes
step-2-16k⚡ 93 tok/sChatStep-2 flagship, 16K context
step-3-chat⚡ 93 tok/sChat🔥 Primary chat model
step-3-reasoning⚡ 85+ tok/sReasoningThinking model
step-3-vl⚡ 93 tok/sVision/OCRVision-language model
step-1.5v-7b⚡ 100+ tok/sChat/VisionSmaller vision model
step-1-flash⚡ 120+ tok/sFast⚡ Ultra-fast lightweight

Audio Models — step-asr (STT) and step-tts (TTS) included in all plans

💻 cURL Example

curl -X POST https://api.stepfun.ai/step_plan/v1/chat/completions \
  -H "Authorization: Bearer 4d2tfWopcoBE9gHPP5boAGiYCrFB7iFiFCt1EluOGoWLVlQVxePCvNh6MwN5I7MKE" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "step-3-chat",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

🐍 Python Example

from openai import OpenAI

client = OpenAI(
    api_key="4d2tfWopcoBE9gHPP5boAGiYCrFB7iFiFCt1EluOGoWLVlQVxePCvNh6MwN5I7MKE",
    base_url="https://api.stepfun.ai/step_plan/v1"
)

response = client.chat.completions.create(
    model="step-3-chat",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

⚠️ Pitfalls & Notes

💡
Own Models Only — StepFun only hosts its own Step model family. No third-party models available.
💡
STT & TTS — StepFun uniquely includes speech-to-text and text-to-speech endpoints alongside chat models.
⚠️
Chinese Optimization — Step models are particularly strong for Chinese/English bilingual tasks.

🏷️ Categories

Chat Vision STT/TTS Reasoning OCR