Get the latest setups
Models/Swift Qwen3.8 27b
Text / ukisai

Swift Qwen3.8 27b

Swift Qwen3.8 27b is a fine-tuned derivative of Qwen/Qwen3.8-27B, released by ukisai under the swift-open-license-1.0. It is an image-text-to-text model with safetensors weights and reasoning tags. According to its official model card, it is described as a reasoning-efficient variant that uses 58.3% fewer thinking tokens while maintaining near-identical performance (<1% loss) and achieving a x1.95 speed-up on several tasks. The training approach involved identifying reasoning-marker tokens that appear to trigger overthinking in Qwen's reasoning rollouts, then fine-tuning by penalizing those tokens while reasoning. It also includes a transfer component derived from BottleCap AI's ThinkingCap-Qwen3.6-27B. Evaluation results are stated to compare the Qwen3.8-27B BF16 base with the same base plus the Swift adapter.

EstimatedRepository opened Sep 8, 2026Source checked 9/24/2026Version: 6b5cb485
LOCALRENTED GPUOPEN WEIGHTS
Freshness 79 out of 100. Repository age: 62 days old, Sep 8, 2026. Measures age, not quality.
Popularity 50 out of 100. 2,460 downloads in 30 days; 639 lifetime likes. Checked Nov 8, 2026. Measures adoption, not quality.
What do these scores mean?

Two ways to compare models. Neither score measures answer quality, benchmark performance, or whether a model fits your machine.

Freshness · release age

A new release starts at 100. The score halves every 180 days: 100 × 2^(−age in days / 180). When the release date is unavailable, a repository date is a fallback—not proof of when the model was released. Missing dates stay unscored.

Check the date source
Popularity · adoption

Downloads over the past 30 days carry 80% of the score; lifetime likes carry 20%. Both use a logarithmic scale, capped at 10 million downloads and 100,000 likes. If likes are unavailable, downloads carry all the weight. Missing download statistics stay unscored.

round(80 × min(1, log10(1 + downloads) / 7) + 20 × min(1, log10(1 + likes) / 5))

Check the statistics source
Scores as of Nov 9, 2026. Statistics checked Nov 8, 2026. Statistics are marked for refresh after 72 hours.Full scoring method

At a glance

Parameters
27.78B
Architecture
qwen3_5
License
swift-open-license-1.0
Disk space
51.7 GB
Software
SGLang, vLLM
View the model source
MY HARDWARE

Will it run on your machine?

Save your machine to see a personalized rating and its reasoning.

Add my hardware

Best for

  • Image-text-to-text tasks where shorter reasoning traces are desired.
  • Reasoning tasks where fewer thinking tokens may reduce computational cost.
  • Scenarios requiring a fine-tuned variant of Qwen3.8-27B with token-efficient reasoning.
  • Applications where the model card's reported x1.95 speed-up on several tasks is relevant.
  • Research or experimentation with reasoning-efficient derivatives of Qwen3.8-27B.

Tradeoffs

  • The model is gated; access is controlled.
  • It uses a custom license (swift-open-license-1.0), which may have specific terms.
  • The official model card claims evaluation only against the Qwen3.8-27B BF16 base with the Swift adapter; no broader independent testing is provided.
  • The reported performance metrics (<1% loss, x1.95 speed-up) are from the model card's stated testing and have not been independently verified here.
  • No specific hardware requirements, compatibility details, or purchase recommendations are provided or inferred.

Ways to run it

SGLang · Linux

Repository-specific command in publisher documentation; confirm dependencies and hardware in the source.

python -m sglang.launch_server --model-path ukisai/Swift-Qwen3.8-27b --dtype bfloat16 --tp-size 1 --context-length 262144 --reasoning-parser qwen3 --tool-call-parser qwen3_coder --port 8000
Official instructions

vLLM · Linux

Repository-specific command in publisher documentation; confirm dependencies and hardware in the source.

vllm serve ukisai/Swift-Qwen3.8-27b --dtype bfloat16 --tensor-parallel-size 1 --max-model-len 262144 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder --port 8000
Official instructions

Explore its uses

SAFETENSORSREASONINGIMAGE-TEXT-TO-TEXT

Keep exploring

Text
Prism ML

Bonsai 2 · 27B

A compressed 27B-class reasoning model with publisher-provided GGUF packs and a dedicated llama.cpp fork for CUDA, Metal, and CPU.

Freshness 82 out of 100. Release age: 53 days old, Sep 17, 2026. Measures age, not quality.
Popularity 85 out of 100. 1,688,388 downloads in 30 days; 2,960 lifetime likes. Checked Nov 8, 2026. Measures adoption, not quality.
MODEL SIZE27.36B
LOCALRENTED GPU
Released Sep 17, 2026
Text
DeepSeek

DeepSeek V4.1 Flash

An open-weight model for reasoning, coding, agents, and image understanding. DeepSeek V4.1 Flash combines a 552B backbone with sparse activation and up to one million tokens of context. Choose the original checkpoint or a community uncensored variant below.

Freshness 79 out of 100. Repository age: 60 days old, Sep 10, 2026. Measures age, not quality.
Popularity 86 out of 100. 1,842,868 downloads in 30 days; 4,898 lifetime likes. Checked Nov 8, 2026. Measures adoption, not quality.
MODEL SIZE552B backbone · 8B active prefill / 16B active decode
LOCALRENTED GPU
Repository opened Sep 10, 2026
Text
Qwen

Qwen3.8

Explore 4 versions of Qwen3.8. Compare model sizes, packages, hardware requirements, and available setups.

4 versionsCompare sizes & formats
LOCALRENTED GPU
Compare versions →