ModelScope

ModelScope

unknown

ModelScope (魔搭) API-Inference serves selected open models through an OpenAI-compatible API; calls are redeemed with free daily login credits (魔粒), and a linked, real-name-verified Alibaba Cloud account is required.

https://api-inference.modelscope.cn/v1

Last Probe Latency

—

Healthy Observed Days

—

Rate Limits

— RPM / — RPD

Sign-up Required

Yes

Provider status summarizes sampled models checked with the monitoring account. Your access and limits may differ.

Documentation checked: . Model availability is subject to the provider's current catalog.

Requires a ModelScope account linked to a real-name-verified Alibaba Cloud account; using credits also requires a verified profile email. Logging in each day grants 200 short-term credits, plus 50 more once an Alibaba Cloud account is linked; short-term credits expire after 24 hours. API-Inference calls cost about 0.5, 1 or 2 credits for lightweight, mainstream or flagship models, and calls are allowed while the credit balance lasts. Credits are shared with AIGC and training services, so this is not a fixed requests-per-day allowance. Concurrency is adjusted dynamically, aiming to keep single-concurrency use working. The service is free and non-commercial, intended for development and evaluation, and should not be used for workloads needing high concurrency or an SLA.

This selection includes three current models whose IDs appear in the public /v1/models list as of 2026-11-19. Older models may be withdrawn from API-Inference as new ones launch. Authenticated inference has not been checked.

Info

Base URLhttps://api-inference.modelscope.cn/v1
Sign-up RequiredYes
Credit CardNot verified
Context Window—
Last Verified2026-11-19
openai-compatiblechina-availablevisionmulti-model

Models(3)

Model health is unknown until an actual check completes.

Qwen/Qwen3.5-35B-A3B

Context: Not published

unknown

Qwen/Qwen3.5-27B

Context: Not published

unknown

ZhipuAI/GLM-4.7-Flash

Context: Not published

unknown

Quick Start

bash
export API_KEY="your_api_key_here"
curl 'https://api-inference.modelscope.cn/v1/chat/completions' \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $API_KEY" \
  -d '{
    "model": "Qwen/Qwen3.5-35B-A3B",
    "messages": [
      {
        "role": "user",
        "content": "Hello! How are you?"
      }
    ]
  }'