Qwen3.8 27B
Qwen (Alibaba) · #9 in our ranking · Best to self-host
Qwen3.8 27B is an open-weight dense model from Qwen (Alibaba), released Aug 2026, with 27B parameters, a 256K-token context window, and the Apache 2.0.
Qwen3.8 27B brings the Qwen3.8 generation to a deployment-friendly dense model. It understands images and video, thinking can be switched off per request, and the Apache 2.0 license makes it simple to run in your own cloud.
Qwen3.8 27B specs
- Provider
- Qwen (Alibaba)
- Architecture
- Dense
- Parameters
- 27B total
- Context
- 256K tokens | 197K words
- License
- Apache 2.0 Commercial use
- Input
- Text, Image, Video
- Released
- Aug 2026
- Weights
- Hugging Face
Sourced from the model card on Hugging Face. Last checked 2026-10-01.
Highlights
- Apache 2.0 license
- Small enough to self-host
- Image and video input
Where it fits for payment teams
- Payment reconciliation
#3 pick. Small enough to run on your own GPUs under Apache 2.0 when ledger data cannot leave your VPC.
- KYC and KYB checks
#1 pick. A compact vision-language model under Apache 2.0, easy to host next to your KYC data.
Harnesses in Runtime
Run Qwen3.8 27B through any of these harnesses on Runtime.
How to run Qwen3.8 27B
Download the weights
Pull Qwen/Qwen3.8-27B from Hugging Face. Check the Apache 2.0 before you deploy.
Serve it in your cloud
Run it with vLLM or SGLang on your own GPUs, or use a managed provider that hosts the model.
Put it to work in Runtime
Point an agent or a single skill at the model through OpenCode, then compare it to your current model with evals.
Qwen3.8 27B FAQ
What is the context window of Qwen3.8 27B?
+
Qwen3.8 27B supports a 256K-token context window (262,144 tokens), roughly 197K words of English text.
How many parameters does Qwen3.8 27B have?
+
Qwen3.8 27B is a dense model with 27B parameters, all active on every token.
Is Qwen3.8 27B free for commercial use?
+
Yes. Qwen3.8 27B is released under the Apache 2.0 license, which is permissive and allows commercial use and fine-tuning.
What can Qwen3.8 27B take as input?
+
Qwen3.8 27B accepts text, image, video input and generates text.
Is Qwen3.8 27B good for payment reconciliation?
+
Yes. Qwen3.8 27B is our #3 pick for payment reconciliation. Small enough to run on your own GPUs under Apache 2.0 when ledger data cannot leave your VPC.
Is Qwen3.8 27B good for KYC and KYB checks?
+
Yes. Qwen3.8 27B is our #1 pick for KYC and KYB checks. A compact vision-language model under Apache 2.0, easy to host next to your KYC data.
Can I run Qwen3.8 27B in my own cloud with Runtime?
+
Yes. Download the weights from Hugging Face (Qwen/Qwen3.8-27B) and serve them with vLLM or SGLang in your cloud, or use a managed provider that hosts the model. Runtime agents can then use Qwen3.8 27B through a harness like OpenCode, with your data staying in your environment.