Qwen 3.8 27B "Overthinking" Debate: Tests Point to Necessary Cost for Performance
Qwen 3.8 27B sparked "overthinking" criticism after users measured 16K+ default reasoning tokens (@emax relayed the complaints). Around 08-17, multiple users defended the model with local benchmarks, converging on the view that the deep thinking is the necessary cost of approaching much larger models, and users should adjust expectations rather than dismiss the model.
Confirmed
- @sukazu, after comparing against GLM 5.3 and DeepSeek V4 Flash/Pro on the same tasks, argued the model is not truly overthinking despite using more tokens than 3.6.
- @maxwell321 ran it locally (3×3090 + Tesla P40, Unsloth UD-Q8KXL quant), finding extensive thinking surfaces fine details from training and brings performance near Claude Sonnet.
- @AltruisticHeat9531 called 16K+ tokens a necessary sacrifice to rival trillion-parameter models, citing Karpathy's "tokens to think" and SWE-Rebench data.
- @hiImMate tested thinking levels (UDQ4XL) on an open-ended HTML card task: the xhigh tier consumed 23.8K tokens but outperformed DS V4 Flash, ChatGPT and Claude Opus 5.
- @ResearchCrafty1804 noted closed-source models (Opus 5, GPT-5.6 Sol) hide internal compute, making raw token comparisons unfair to open models.
- @wombweed advised disabling thinking or lowering the reasoning budget when speed matters, and trying higher quantization if loops appear; heavy thinking is the product of RL training.
- @Afinetheorem rated it one of the best local LLMs for high-end consumer GPUs, excellent at coding but among the worst tested at logic and visual reasoning (46/58 vs SOTA).
Why it matters
- The debate exposes an evaluation asymmetry: open models' reasoning costs are measurable and criticizable, while closed models hide compute server-side.
- It offers local-deployment users a tradeoff guide: strong value on coding/generation tasks, but logic and vision workloads call for other models.
2026-08-17 ~ 2026-08-18 · 7 related posts
- Episode 1: Qwen 3.8 27B "Overthinking" Debate: Tests Point to Necessary Cost for Performance(2026-08-17, 7 posts)
- Episode 2: Qwen3.8-27B matches top closed-source models on benchmarks while running on a single consumer GPU(2026-08-17, 20 posts)
- Episode 3: Qwen 3.8 Default Temperature Causes Gibberish, Lower It(2026-08-18, 2 posts)
Primary sources
- [source] Local test: Qwen 3.8 27B's overthinking brings it near Sonnet-level performance — maxwell321 · 2026-08-17
- [source] Why Qwen 3.8 27B Isn't Overthinking: Compared with GLM and DeepSeek — sukazu · 2026-08-17
- Netizen jokes Qwen 3.8 27B defaults to 'overthinking' — emax · 2026-08-17
- Qwen 3.8 27b shines in code but fails logic tests, highlighting local LLM trade-offs — Afinetheorem · 2026-08-17
- Qwen 3.8's High Token Cost is a Fair Trade for Performance — Altruistic_Heat_9531 · 2026-08-17
- Qwen's Overthinking is a Feature, Adjust Your Expectations — wombweed · 2026-08-18
- [source] Qwen3.8 27B xhigh burns 23.8K tokens overthinking a simple card prompt — and crushes every rival — hiImMate · 2026-08-18