Home › Tool Registry › Qwen 3.8
Qwen 3.8
by Alibaba
Visit Website ↗

Note: there is no Alibaba release officially called "Qwen 3.8" — the name most likely refers to the Qwen3 family (Qwen3, Qwen3-Max, Qwen3-Omni) or to Qwen2.5/Qwen2.5-Max, Alibaba's open-weight large language model series. Qwen models are known for strong multilingual and coding performance, permissive licensing on many sizes, and availability both as downloadable weights and via Alibaba Cloud APIs.

2
Total mentions
2
Articles
See website
Pricing
No
Free tier
← Back to registry
Articles mentioning Qwen 3.8 (2)
01Open_Weights_Open_Future_Alibaba_Qwen_3.8_Apache_2.0Aug 14, 2026 02Alibaba_Qwen_3.8_Tops_Kimi_K3_Second_Only_to_Fable_5Jul 19, 2026
Key Features
Dense and Mixture-of-Experts variants from ~0.6B to 235B+ parameters
Native multilingual support (100+ languages) with strong Chinese and English ability
Hybrid reasoning modes and extended thinking in Qwen3 models
Tool/function calling, agentic workflows, and structured output support
Open weights under Apache 2.0 for most sizes, plus hosted APIs via Alibaba Cloud Model Studio and Qwen Chat
Pros & Cons
Pros
Top-tier open-weight performance, often competitive with closed models on coding and math
Permissive Apache 2.0 licensing on most sizes enables commercial self-hosting
Broad deployment support across Hugging Face, Ollama, vLLM, llama.cpp and cloud providers
Cons
Version naming is confusing (Qwen/Qwen1.5/Qwen2/Qwen2.5/Qwen3 and Max/Omni/Turbo suffixes)
Largest and newest flagship models are API-only, not open-weight
Documentation quality and English-language evaluation coverage lag some Western competitors