Conversation / Quote
Qwen3.8-2.4T-A95B(Open-source) API Price
Alibaba Bailian · qwen3.8-2.4t-a95b
| Billing | CNY · 1M tokens |
|---|---|
| Context Window | 1000K tokens |
| Max Output | 131K tokens |
| Modalities | Text |
| Deep Thinking | Supported |
| Verified Date | 2026-09-07 |
| Official pricing page | help.aliyun.com/zh/model-studio/qwen3-8-2-4t-a95b ↗ |
Source text: the note below is the model's official text, kept in its original Chinese and not translated.
2026-08 发布的开源旗舰 MoE(2.4T 总参/激活 95B,混合注意力,1M 上下文);刊例价与 qwen3.8-max 相同(缓存命中 1.5/显式缓存创建 15/命中 1)。
Report a pricing issue
Found inaccurate pricing or model information? We will review it.
Monthly Cost Estimate
30 days · no cache hits · list price
Free quotas, tiered pricing, batch discounts, and time-based pricing are not included; refer to official bills.
Alibaba Bailian Related Models
- Qwen3.8-Max-Prime(Prime mode) ¥24 / ¥72
- Qwen3.8-Max ¥12 / ¥36
- Qwen3.8-Max(v0902) ¥12 / ¥36
- Qwen3.8-27B(Open-source) ¥3 / ¥12
- Qwen3-Max ¥2.5 / ¥10
- Qwen3-Max(v2026-01-23) ¥2.5 / ¥10
- Qwen3-Max(v2025-09-23) ¥6 / ¥24
- Qwen3.8-Flash ¥0.8 / ¥2.7
- Qwen3.7-Flash ¥0.2 / ¥0.8
- Qwen-Max ¥2.4 / ¥9.6
- Qwen-Plus ¥0.8 / ¥2
- Qwen-Plus-Latest ¥0.8 / ¥2
- Qwen-Flash ¥0.15 / ¥1.5
- Qwen-Turbo ¥0.3 / ¥0.6
- Qwen-Long ¥0.5 / ¥2
- Qwen-Math-Plus ¥4 / ¥12
- Qwen-Math-Turbo ¥2 / ¥6
- Qwen-MT-Plus ¥1.8 / ¥5.4
- Qwen-MT-Turbo ¥0.7 / ¥1.95
- Qwen-MT-Flash ¥0.7 / ¥1.95
- Qwen-MT-Lite ¥0.6 / ¥1.6
- Qwen-MT-Uni ¥20 /1M tokens
- Qwen-Deep-Research ¥54 / ¥163
- Decision Model Preview Free
- Qwen-VL-Max ¥1.6 / ¥4
- Qwen-VL-Plus ¥0.8 / ¥2
- Qwen-VL-OCR ¥0.3 / ¥0.5
- Qwen3.5-OCR ¥0.5 / ¥2
- Qwen-Omni-Turbo ¥0.4 / ¥1.6
- QwQ-Plus ¥1.6 / ¥4
- QVQ-Max ¥8 / ¥32
- Qwen3-8B ¥0.5 / ¥2
- Qwen3-14B ¥1 / ¥4
- Qwen3-32B ¥2 / ¥8
- Qwen3-30B-A3B(Open-source) ¥0.75 / ¥3
- Qwen3-30B-A3B-Instruct(v2507) ¥0.75 / ¥3
- Qwen3-30B-A3B-Thinking(v2507) ¥0.75 / ¥7.5
- Qwen3-235B-A22B ¥2 / ¥8
- Qwen3-235B-A22B-Instruct(v2507) ¥2 / ¥8
- Qwen3-235B-A22B-Thinking(v2507) ¥2 / ¥20
- Qwen3-Next-80B-A3B-Instruct ¥1 / ¥4
- Qwen3-Next-80B-A3B-Thinking ¥1 / ¥10
- Qwen3-Coder-480B-A35B-Instruct ¥6 / ¥24
- Qwen3-Coder-Flash ¥1 / ¥4
- Qwen3-Coder-Plus ¥4 / ¥16
- Qwen-Coder-Plus ¥3.5 / ¥7
- Qwen3-Omni-Flash ¥1.8 / ¥6.9
- Qwen3-Omni-Flash-Realtime ¥2.2 / ¥8.3
- Qwen3.8-Omni-Flash ¥0.8 / ¥2.7
- Qwen3.8-Omni-Flash-Realtime ¥1.5 / ¥4.5
- Qwen-Audio-3.1-Realtime-Plus ¥5 / ¥40
- Qwen3.5-Omni-Plus ¥7 / ¥40
- Qwen3.5-Omni-Flash ¥2.2 / ¥13.3
- Qwen3.5-Omni-Plus-Realtime ¥10 / ¥60
- Qwen3.5-Omni-Flash-Realtime ¥3.3 / ¥20
- Qwen3-VL-Plus ¥1 / ¥10
- Qwen3-VL-Flash ¥0.15 / ¥1.5
- Qwen3-LiveTranslate-Flash-Realtime ¥64 / ¥64
- Qwen3.5-LiveTranslate-Flash-Realtime ¥40 / ¥100
- Qwen3.8-LiveTranslate-Flash-Realtime ¥40 / ¥100
- Qwen3.5-Flash ¥0.2 / ¥2
- Qwen3.5-Plus ¥0.8 / ¥4.8
- Qwen3.5-Plus(v2026-04-20) ¥0.8 / ¥4.8
- Qwen3.5-27B ¥0.6 / ¥4.8
- Qwen3.5-35B-A3B ¥0.4 / ¥3.2
- Qwen3.5-122B-A10B ¥0.8 / ¥6.4
- Qwen3.5-397B-A17B ¥1.2 / ¥7.2
- Qwen3.6-Flash ¥1.2 / ¥7.2
- Qwen3.6-Plus ¥2 / ¥12
- Qwen3.6-27B ¥3 / ¥18
- Qwen3.6-35B-A3B ¥1.8 / ¥10.8
- Qwen3.6-Max-Preview ¥9 / ¥54
- Qwen3.7-Max ¥12 / ¥36
- Qwen3.7-Max(v2026-06-08) ¥12 / ¥36
- Qwen3.7-Plus ¥2 / ¥8
- Tongyi Wanx wan2.6-t2i ¥0.2 /image
- Tongyi Wanx wan2.5-t2i-preview ¥0.2 /image
- Tongyi Wanx wan2.2-t2i-plus ¥0.2 /image
- Tongyi Wanx wan2.2-t2i-flash ¥0.14 /image
- Qwen-Image 3.0 Pro ¥0.25 /image
- Qwen-Image 3.0 ¥0.18 /image
- Z-Image Turbo ¥0.1 /image
- Wanx 2.7 Image Generation & Editing Pro ¥0.5 /image
- Qwen-MT-Image 2.0 ¥0.004 /image
- Wanx 2.7 Image Generation & Editing ¥0.2 /image
- Wanx 2.7 Text-to-Video ¥0.6 /sec
- Wanx 2.5 Text-to-Video Preview ¥0.3 /sec
- Wanx 2.6 Text-to-Video ¥0.6 /sec
- Wanx 2.1 Text-to-Video Turbo ¥0.24 /sec
- Wanx 2.1 Text-to-Video Plus ¥0.7 /sec
- Wanx 3.0 Video ¥0.3 /sec
- Wanx 3.0 Video Prime ¥0.45 /sec
- HappyOyster World Exploration (Adventure) ¥0.05 /sec
- HappyOyster Realtime Directing ¥2 /sec
- Qwen-Audio-3.1-TTS-Next ¥6 / ¥12
- Qwen-Audio-3.1-ASR-Flash ¥0.8 / ¥2.7
- Qwen-Audio-3.1-ASR-Flash-Filetrans ¥0.8 / ¥2.7
- Qwen-Audio-3.1-ASR-Flash-Streaming ¥6 / ¥4.5
- Fun-Music Preview ¥0.005 /sec
- Fun-Music V1 ¥0.002 /sec
FAQ
How much does the Qwen3.8-2.4T-A95B(Open-source) API cost?
Input ¥12/1M tokens, output ¥36/1M tokens
How large is the Qwen3.8-2.4T-A95B(Open-source) context window?
1000K tokens, up to 131K tokens of output. Handles long documents, multi-turn conversations and large-scale inputs.
Which input modalities does Qwen3.8-2.4T-A95B(Open-source) support?
Text. Supports deep thinking / reasoning.
What is the approximate monthly cost of Qwen3.8-2.4T-A95B(Open-source)?
Depends on volume: light use (1M input + 300K output per day) ≈ ¥684/month; medium (5M + 1.5M) ≈ ¥3,420/month; heavy agents (10M + 3M) ≈ ¥6,840/month. Cache hits, batch discounts and tiered pricing can lower this further.
How does Qwen3.8-2.4T-A95B(Open-source) compare with peers on value?
Alibaba Bailian Qwen3.8-2.4T-A95B(Open-source) is competitively priced among comparable models; compare against similar models. See the Pricemon market table for a full comparison.
What use cases suit Qwen3.8-2.4T-A95B(Open-source)?
Suited to applications in its domain.The long context supports deep-analysis scenarios.
Does Qwen3.8-2.4T-A95B(Open-source) have a free tier?
Check Alibaba Bailian's official pricing page for the latest free quota, trial policy and tiered pricing. Policies differ by vendor and change over time.