Qwen/Qwen3.8-27B

Timeline

1. Oct 11, 2026 · Opinion / analysis · 1 source
   UPDATE: Qwen 3.8 27B 140 tok/s on single RTX 3090 Megakernel: KL divergence 0.0009 vs llama.cpp
   RECAP: The megakernel is a CUDA engine for Qwen3.8-27B that runs 1.4-1.9x faster than llama.cpp on a single 3090.
2. Oct 11, 2026 · Opinion / analysis · 1 source
   Now you can grow Bonsai on your potato
   Full 27B-class reasoning in ternary transformer weights, for llama.cpp (CUDA, Metal, CPU)
3. Oct 9, 2026 · Model release · 2 sources
   ConwayResearch/Underdog-Saluki-27B-1.0
   ConwayResearch published the model Underdog-Saluki-27B-1.0 on Hugging Face.
4. Oct 9, 2026 · Benchmark result · 1 source
   Qwen3.8 27B Xhigh or Flash Next Medium
   In my Qwen3.8 27B reasoning benchmarks, Xhigh generally produced the strongest results, but it needed much longer responses than Medium.
5. Oct 6, 2026 · Benchmark result · 1 source
   Qwen3.8 Flash Next Reasoning Modes: Off vs Low vs Medium vs Xhigh
   Like Qwen3.8 27B, Qwen3.8 Flash Next has three reasoning efforts: low, medium, and xhigh.
6. Oct 5, 2026 · Model release · 1 source
   perplexity-ai/pplx-decider-v1.1-27b
   perplexity-ai published the model pplx-decider-v1.1-27b on Hugging Face.
7. Oct 4, 2026 · Tutorial / explainer · 1 source
   Qwen3.8 27B addition in words
   Research: Qwen3.8 27B addition in words
8. Oct 4, 2026 · Benchmark result · 1 source
   Which is The Best Qwen3.8-27B?
   Which is the best fine-tune of Qwen3.8 27B that reduces the amount of thinking and reasoning tokens but still keeps the best accuracy for your particular use case?
9. Oct 3, 2026 · Opinion / analysis · 1 source
   Qwen3.8 Flash Next, Swift 1.5, and MiMo V2.6: What I’m Testing Now
   The Kaitchup – AI on a Budget is a reader-supported publication.
10. Sep 30, 2026 · Benchmark result · 1 source
    Qwen3.8 Flash Next GGUF Benchmark: Q4 to Q1 Accuracy and Token Efficiency
    Quantizing Qwen3.8 Flash Next is almost unavoidable if you want to run it locally.
11. Sep 28, 2026 · Model release · 1 source
    Holo4: powering generalist computer-use agents
    Holo4 is our new series of agentic models.

Often appears with