Qwen3.8-27B
27B native multimodal dense, with an official 3.7-Plus comparison table
Qwen3.8-27B is a 27B native multimodal dense model (model card updated 2026-08-14). The official model card lists a native 262,144-token context, extensible to 1,000,000 with YaRN, Apache 2.0, and weights at Hugging Face Qwen/Qwen3.8-27B. The official hosted API also lists qwen3.8-27b: 1M context, official price $0.5 input / $3 output per 1M tokens. QCode lists qwen3.7-plus and qwen3.8-max; whether the 27B is available is decided by /models.
Updated 2026-10-22
One key across Claude, GPT and Chinese models, billed per token.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Local-first facts
Dense multimodal
Native multimodal dense model, not the 2.4T MoE. Official @Alibaba_Qwen.
Native context
Native 262K. 1M needs YaRN. It is not a native 1M window.
YaRN stretch
Official post: YaRN can take it to 1M. Do not write that as native.
License
Open weights. HF path Qwen/Qwen3.8-27B.
What the 27B is
Qwen3.8-27B is open-weight, with the model card updated on 2026-08-14: a 27B native multimodal dense model for local or self-hosted clusters. The official model card lists a native 262,144-token context, YaRN up to 1,000,000, Apache 2.0, and the repo Hugging Face Qwen/Qwen3.8-27B. It also has an official hosted version, qwen3.8-27b, with 1M context and an official price of $0.5 / $3 per 1M tokens. It is not the hosted Max.
From open weights to hosted API: how to read the 3.7-Plus comparison
Official model card: 27B dense, 262K native, YaRN to 1M, Apache 2.0. Official hosted page: qwen3.8-27b, 1M context, max input 991K, max output 131K, official price $0.5 input / $3 output per 1M. The social-media claim of beating Qwen3.7-Plus overall is a vendor statement; no independent leaderboard is built here.
Timeline
3.8-Max ships and open weights are promised.
@Alibaba_Qwen drops Qwen3.8-27B and Qwen3.8-2.4T-A95B.
Weights are downloadable. QCode lists qwen3.8-max and qwen3.7-plus; whether the 27B is available is decided by /models.
Confirmed vs treat carefully
Confirmed
Official model card (updated 2026-08-14): 27B native multimodal dense; native 262,144 tokens, YaRN to 1,000,000; Apache 2.0; HF Qwen/Qwen3.8-27B. Official hosted page: qwen3.8-27b, 1M context, max input 991K, max output 131K, official price $0.5 input / $3 output per 1M tokens.
Treat carefully
"Native 1M" is inaccurate: the model card says 262K natively, and 1M needs YaRN. The social-media claim of beating 3.7-Plus overall is no longer this page's basis; the official model card provides a comparison table. QCode lists qwen3.7-plus and qwen3.8-max; whether the 27B is available is decided by /models.
Local vs API
Run the 27B yourself
Want weights, a clean license, and your own runtime? Pull Qwen/Qwen3.8-27B. Apache 2.0.
Call Qwen on QCode
QCode lists qwen3.7-plus and qwen3.8-max; whether qwen3.8-27B is available is decided by /models.
How to run it locally
Pull the weights from Hugging Face Qwen/Qwen3.8-27B (or the matching ModelScope repo). Configure the native 262,144-token context. For 1,000,000, enable YaRN per the official instructions; do not assume every framework has it on by default. That is the self-hosted path. For hosted use, the official page lists qwen3.8-27b, 1M context and an official price of $0.5 / $3 per 1M tokens.
On QCode
QCode /models lists qwen3.8-max and qwen3.7-plus, both callable; whether qwen3.8-27B is available is decided by /models. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; with an active Personal Plan you cannot buy a pack, so buy a booster pack when the plan quota is short.
Qwen3.8-27B FAQ
What is the 27B?
A 27B native multimodal dense model. The official model card was updated on 2026-08-14; weights are at Hugging Face Qwen/Qwen3.8-27B.
Does it beat 3.7-Plus?
The official model card provides a comparison table against Qwen3.7-Plus; the social-media "beats overall" claim is not treated as a conclusion here, and no independent score is built.
How long is context?
Native 262K. YaRN can go to 1M — that is not the native window.
License?
Apache 2.0. HF: Qwen/Qwen3.8-27B.
Can I call the 27B on QCode?
QCode currently lists qwen3.8-max and qwen3.7-plus as callable; whether qwen3.8-27B is available is decided by /models.
Which API id then?
On QCode, use qwen3.8-max or qwen3.7-plus for Qwen models; availability is as shown on /models.
Sources
Official Hugging Face model card Qwen/Qwen3.8-27B (updated 2026-08-14, with the 3.7-Plus comparison table), the QwenCloud qwen3.8-27b model page and QCode /models, both captured 2026-10-20.
Run 27B locally, call the 3.7-tier qwen3.7-plus or qwen3.8-max
27B is open-weight; QCode lists qwen3.8-max and qwen3.7-plus, with availability on /models.
Related
Qwen 4 Release Date: official status (2026-11-05)
Alibaba has not announced a Qwen 4 release date: the press release wording, what GitHub and the model overview show, and community speculation with its source; plus Qwen3.8 on QCode.
The Complete Guide to Qwen3.7-Plus
On sale alongside Max, real usage data verified
2026 model radar
Which ids are actually in the catalog.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Not affiliated with Alibaba / Qwen. Qwen3.8-27B is open-weight, with the 3.7-Plus comparison in the official model card; QCode lists qwen3.8-max and qwen3.7-plus, and whether the 27B is available is decided by /models. Checked 2026-10-20.