Skip to content
Computer vision

Qwen3.8-27B

компактная dense-модель с нативным зрением и длинным контекстом

1 min read

Qwen3.8-27B — модель семейства Qwen на 27B параметров: causal LM с vision-энкодером. Это dense-модель для кода, профессиональных задач, исследований и длительных agentic-сценариев; понимает изображения и видео и поддерживает управление режимом размышления.

Ключевые особенности

  • Улучшения в coding, professional work, research и long-horizon agentic tasks.
  • Нативное понимание изображений и видео: от STEM-схем и документов до часовых видео.
  • Гибкое управление thinking: режим включён по умолчанию, его можно отключить; глубина задаётся через reasoning_effort, а reasoning-контекст сохраняется через preserve_thinking.
  • Контекст 262 144 токена нативно, расширяется до 1 000 000.
  • Совместимость с Hugging Face Transformers, vLLM, SGLang и TokenSpeed.

Пример

Загрузите в модель часовое видео или пакет документов и попросите составить пошаговый план действий: Qwen3.8-27B рассчитан на сложные многошаговые задачи с обработкой обратной связи от среды.

ПодробнееБенчмарки: Terminal Bench 2.1 — 73.0, SWE-bench Pro — 61.7, QwenSWEBench — 79.0.
Архитектура: 64 слоя, hidden 5120, FFN 17 408, MTP-обучение.
Требования к VRAM и железу в карточке не указаны.
Hosted-вариант в Qwen Cloud анонсирован, но сервис ещё не запущен.
Discuss Telegram 0 27

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # Uhvg856/Qwen3.8-27
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download Uhvg856/Qwen3.8-27 --local-dir ~/models/Qwen3.8-27
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to Qwen3.8-27
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of Qwen3.8-27 (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

Uhvg856/Qwen3.8-27
Author
Uhvg856
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

270

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments