Qwen3-8B
Alibaba Cloud (Qwen) · China
Based on published licence terms, Qwen3-8B is released under standard Apache 2.0 with no field-of-use carve-outs, making self-hosted commercial deployment viable. Training-data disclosure is limited to a token count and Chinese origin creates EU AI Act Art. 53 transparency and data-transfer risks that deployers should document.
Licence facts
- Parameters
- 8.2B total (~6.95B non-embedding), 36 layers
- Architecture
- Dense transformer with Grouped-Query Attention (32 Q heads / 8 KV heads), dual thinking / non-thinking inference modes
- Context length
- 32K native, 131K via YaRN scaling
- Languages
- Multilingual — Qwen 3 series covers ~119 languages
- Released
- 2025-04-29
Known risks
- Jurisdictional exposure: if inference is served via Alibaba's Chinese regions or Model Studio, PRC data-access statutes (DSL, PIPL, National Intelligence Law) may conflict with GDPR Chapter V transfer requirements. Self-hosting in the EEA or a non-PRC provider largely neutralises this for the weights themselves.
- Training-data transparency gap: the Qwen team discloses a headline ~36T tokens across the Qwen 3 series but no dataset enumeration, source breakdown, opt-out mechanism, or copyrighted-content filtering — thin for Art. 53(1)(d) GPAI training-data summary obligations.
- Operational controls if using Alibaba Cloud EU: Alibaba Cloud offers a Frankfurt region, an EEA DPA, SCCs, and adherence to the EU Cloud Code of Conduct, but sub-processor chains and support-access paths should be audited, and parent-entity controls warrant documented assessment.
Sources
See also
Reviewed by Ali Madjaji · Last reviewed 2026-04-17· Reviewed 1 day agoSuggest a correction