What's free
~2,000 free API calls/day across open-weight models (Qwen3, DeepSeek, GLM, Llama, etc.) via API-Inference
Rate limits
~2,000 calls/day; concurrency/QPS caps applied and dynamically adjusted
The catch
Alibaba's model hub — a different product from Alibaba Model Studio. Requires a ModelScope account bound to an Alibaba Cloud account with real-name (ID) verification — a practical barrier for non-China users. Explicitly non-commercial ("for developers to experience"). No card.
Free models · sample
LLM-Research/Llama-4-Maverick-17B-128E-InstructMedAIBase/AntAngelMedMiniMax/MiniMax-M1-80kMusePublic/Qwen-Image-EditOpenGVLab/InternVL3_5-241B-A28BPaddlePaddle/ERNIE-4.5-0.3B-PTQwen/Qwen-Image-EditShanghai_AI_Laboratory/Intern-S1A sample of models reachable on the free tier — the live catalog changes. Pull the current set with GET https://api-inference.modelscope.cn/v1/models.
Quickstart — chat completions
from openai import OpenAI
client = OpenAI(base_url="https://api-inference.modelscope.cn/v1", api_key="<YOUR_FREE_API_KEY>")
resp = client.chat.completions.create(
model="LLM-Research/Llama-4-Maverick-17B-128E-Instruct",
messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)…or with curl:
curl https://api-inference.modelscope.cn/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"LLM-Research/Llama-4-Maverick-17B-128E-Instruct","messages":[{"role":"user","content":"Hello!"}]}'