# ModelScope

Rate-limited free · API key · code · Last updated: 08 Nov 2026

7 free models on ModelScope including Qwen3 30B A3B Thinking 2507, Qwen3 235B A22B Instruct 2507, and Qwen3 30B A3B Instruct 2507. OpenAI-compatible endpoint (https://api-inference.modelscope.cn/v1), up to 256K context. Last updated: 08 Nov 2026.

The Grant: ModelScope (Alibaba's open-source model community) offers free hosted inference for community and Qwen open-weights models.
The Output: Standard OpenAI-compatible completion endpoints with 262K token context windows.
The Process: Create an account on ModelScope and retrieve an API access token.
The Condition: Rate-limited to community usage tiers with fair-share queueing.

Source: https://modelscope.cn/docs/api-inference · Last updated: 08 Nov 2026

| Model | Cost | Context | RPM | TPM | RPD | Day tokens | Verified |
| --- | --- | --- | --- | --- | --- | --- | --- |
| Qwen3 30B A3B Thinking 2507 | $0 | 256K | 30 | Not published | 500 | Not published | 2026-11-08 |
| Qwen3 235B A22B Instruct 2507 | $0 | 256K | 30 | Not published | 500 | Not published | 2026-11-08 |
| Qwen3 30B A3B Instruct 2507 | $0 | 256K | 30 | Not published | 500 | Not published | 2026-11-08 |
| Qwen3-235B-A22B-Thinking-2507 | $0 | 256K | 30 | Not published | 500 | Not published | 2026-11-08 |
| Qwen3 Coder 30B A3B Instruct | $0 | 256K | 30 | Not published | 500 | Not published | 2026-11-08 |
| GLM-4.5 | $0 | 128K | 30 | Not published | 500 | Not published | 2026-11-08 |
| GLM-4.6 | $0 | 198K | 30 | Not published | 500 | Not published | 2026-11-08 |

Machine-readable: /data/modelscope.json

Numbers cross-checked against official docs and third-party trackers as of 2026-11-08. Limits change without notice — this is a map, not a contract.