Alibaba Cloud Model Studio
International (Singapore) Qwen API access via Model Studio with a per-model free token grant, OpenAI-compatible.
- Tokens / month
- 1M
- Credit window
- 90 days
- Card required
- No
Alibaba Cloud Model Studio scores 4.7/10 for free value in LLM. The main constraint: Each supported model includes 1,000,000 free tokens, valid for 90 days after activating Model Studio (confirmed on official model-list page). Covers flagship Qwen3.x-Plus/Flash (up to 1M-token context) and many others; OpenAI-compatible endpoint.
Free value score
4.7 / 10Show the math
Key metrics
Models
39 free · 200+ totalLimits may be stale — last re-verified outside this section's freshness window.
| Model | Family | AA | Access |
|---|---|---|---|
| Qwen3.5-27B | Qwen | 33.8 | free |
| Qwen3.5-122B-A10B | Qwen | 32.3 | free |
| Qwen3.5-397B-A17B | Qwen | 32 | free |
| Qwen3.5-35B-A3B | Qwen | 29.3 | free |
| Qwen3-Max (qwen3-max-2025-09-23, non-thinking) | Qwen | 24 | free |
| Qwen3-Max (qwen3-max-2026-01-23, Thinking) | Qwen | 24 | free |
| CosyVoice (v2 / v3-flash/plus / v3.5) | Qwen | — | free |
| Fun-ASR / Paraformer | Qwen | — | free |
| multimodal-embedding-v1 | Qwen | — | free |
| QVQ (visual reasoning) | Qwen | — | free |
| Qwen-Flash (qwen-flash-2025-07-28) | Qwen | — | free |
| Qwen-Image / Qwen-Image-2.0(-Pro) / Qwen-Image-Max | Qwen | — | free |
| Qwen-Image-Edit (Plus/Max) | Qwen | — | free |
| Qwen-Math-Plus / Qwen-Math-Turbo | Qwen | — | free |
| Qwen-Max (qwen-max-2025-01-25, aka Qwen2.5-Max) | Qwen | — | free |
| Qwen-Omni-Turbo (+ Realtime) | Qwen | — | free |
| Qwen-Plus (qwen-plus-2025-12-01) | Qwen | — | free |
| Qwen-Turbo | Qwen | — | free |
| Qwen-VL-Max (qwen-vl-max-2025-08-13) | Qwen | — | free |
| Qwen-VL-OCR | Qwen | — | free |
| Qwen-VL-Plus | Qwen | — | free |
| Qwen3 dense (32B / 14B / 8B / 4B / 1.7B / 0.6B) | Qwen | — | free |
| Qwen3-235B-A22B-Instruct-2507 | Qwen | — | free |
| Qwen3-235B-A22B-Thinking-2507 | Qwen | — | free |
| Qwen3-30B-A3B (Thinking-2507 / Instruct-2507 / base) | Qwen | — | free |
| Qwen3-ASR-Flash (+ Realtime / Filetrans) | Qwen | — | free |
| Qwen3-Coder-480B-A35B-Instruct | Qwen | — | free |
| Qwen3-Coder-Flash / Qwen3-Coder-30B-A3B-Instruct | Qwen | — | free |
| Qwen3-Coder-Plus (qwen3-coder-plus-2025-09-23) | Qwen | — | free |
| Qwen3-Next-80B-A3B-Thinking / Instruct | Qwen | — | free |
| Qwen3-Omni-Flash (+ Realtime) | Qwen | — | free |
| Qwen3-TTS-Flash (+ Realtime / Instruct / VD) | Qwen | — | free |
| Qwen3-VL (235B-A22B / 30B-A3B / 32B / 8B; Instruct & Thinking) | Qwen | — | free |
| Qwen3.5-Flash (qwen3.5-flash-2026-02-23) | Qwen | — | free |
| Qwen3.5-Plus (qwen3.5-plus-2026-02-15) | Qwen | — | free |
| QwQ-Plus / QwQ-32B | Qwen | — | free |
| text-embedding-v4 / v3 | Qwen | — | free |
| Wan2.6 / Wan2.5 / Wan2.1 (text-to-image, image-edit, text/image-to-video) | Wan | — | free |
| Z-Image | Qwen | — | free |
| Qwen3-LiveTranslate-Flash-Realtime | Qwen | — | paid |
Showing first 39 of a larger catalog.
What you get
1,000,000 free tokens per model (90-day window)
Qwen3.5-Plus / Flash with up to 1M-token context
OpenAI-compatible API (international Singapore region)
Wide model catalog incl. DeepSeek distills, Kimi
Gotchas
- 90-day expiry after activation
Free token grants for each model expire 90 days after you activate Model Studio, regardless of whether they are consumed.
- 1M tokens per model (one-time)
Each supported model carries a one-time grant of 1,000,000 tokens that does not renew once exhausted.
- Alibaba Cloud account required
You must register an Alibaba Cloud account (international or mainland); the free grant activates upon enabling the Model Studio service, not on signup.
Sources
-
"qwen3.7-max ... 1 million tokens for input and output each Valid for 90 days after you activate Model Studio"
https://www.alibabacloud.com/help/en/model-studio/model-pricing -
"The free quota is valid for 30 to 90 days, starting from the date you activate Alibaba Cloud Model Studio ... Starting from 3:00 UTC on September 8, users who activate Alibaba Cloud Model Studio for the first time will receive a free quota that is valid for 90 days."
https://www.alibabacloud.com/help/en/model-studio/new-free-quota -
"Qwen3.7 Max scores 57 on the Artificial Analysis Intelligence Index, placing it well above average among comparable models (averaging 36)."
https://artificialanalysis.ai/models/qwen3-7-max -
"qwen3.7-max | International | 600 | 1,000,000 ... qwen3.5-plus | International | 15,000 | 6,000,000"
https://www.alibabacloud.com/help/en/model-studio/rate-limit -
"Alibaba's Qwen family of AI models has reached 700 million downloads on the Hugging Face platform as of January 2026, making it the world's most [downloaded open-source model family]"
https://www.scmp.com/tech/big-tech/article/3339568/alibabas-qwen-family-hits-700-million-downloads-lead-global-open-source-ai-adoption
Alternatives
Same category · similar quota
4.1B credits/mo for mimo-v2.5 / v2.5-pro (native vision+audio, 1M ctx) — ~550M real tokens/mo on v2.5 or ~200M on Pro in cache-heavy coding; OpenAI + Anthropic endpoints (Claude Code, OpenCode, OpenClaw); 0.8x night discount
