Compact 27B-parameter model from Alibaba's Qwen family, runnable locally on a single high-end consumer GPU.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Qwen 3.8 27B is a mid-sized model in Alibaba's Qwen series, listed here as an open-weights release. The Qwen project publishes its models through qwen.ai, GitHub and Hugging Face, and the Qwen platform covers chat, image and video understanding, document processing, web search and tool use.
According to a YouTube review (Matthew Berman), the 27B model fits on a single RTX 4090 or 5090 and runs at roughly 25-70 tokens per second. The same review gave it an intelligence index of about 34 and an amortized local-hosting cost of about $144 per month. These figures come from a creator's chart, not from the vendor.
The provided pages contain no details on architecture, context length, benchmarks, release date or license. Modalities and license below are therefore not confirmed. Earlier Qwen releases of similar size have mostly used Apache 2.0, but check the Hugging Face model card before relying on that.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
A compact open-weights model that fits on an RTX 4090/5090, running at roughly 25-70 tokens per second. On the cost chart it comes to about $144/month amortized with an intelligence index of about 34.
A 27B open-weight model that runs locally on 24-32 GB machines and ranks well on Artificial Analysis indexes. The creator ran it via LM Studio and Hermes; it failed his Mega Bonk game test and was very slow.
27B dense multimodal open model that beats Opus 4.6 Max on many agentic and coding benchmarks, with 1M context. Full model is 56 GB; Unsloth GGUFs go down to 9 GB.
Closest neighbours by meaning, not just by category.
Model · Alibaba
Alibaba's Qwen 3.5 large open-weights language model (~122B parameters) from the Qwen family.
Model · Alibaba
Small 0.8B-parameter open-weights language model in Alibaba's Qwen 3.5 family, aimed at lightweight and on-device use.
Model · Alibaba
Small 0.5B-parameter open-weights base language model from Alibaba's Qwen2.5 series, Apache 2.0 licensed.
Model · Alibaba
Alibaba's open-weights 2.4T-parameter MoE language model (95B active) from the Qwen3.8 family, with a hosted Max version.
Model · Alibaba
Alibaba's Qwen3.8 Flash Next: a 125B-parameter MoE language model with about 6B active parameters, previewing Qwen 4 design.
Model · Alibaba
Alibaba's Qwen coding-specialist language model, available as open weights for local and hosted use.
