# Qwen 3.6 27B GGUF quality benchmark per-quant KL (localbench, 87 quants) > Uploaders and counts: unsloth 21, lmstudio-community 3, Jackrong 11, bartowski 26, mradermacher (i1) 23, ubergarm 2 (ik_llama.cpp), ggml-org 1. Parent: [Mac local LLMs: Quantization evaluation](https://llms-explorer.com/tree/mac-local-llms-quantization-evaluation/) · 1 facets · 16 facts · page: https://llms-explorer.com/tree/qwen-3-6-27b-gguf-quality-benchmark-per-quant-kl/ ## Facts - Uploaders and counts: unsloth 21, lmstudio-community 3, Jackrong 11, bartowski 26, mradermacher (i1) 23, ubergarm 2 (ik_llama.cpp), ggml-org 1. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - Method is the shared localbench one: about 250,000 tokens in six categories, top-40 KL on prompt tokens through the model's chat template. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - The frontier table lists, for each size, the lowest-KL quant; a quant absent from it is dominated by a smaller file with lower KL. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - 25 Apr 2026: post published by oobabooga, after the 7 Apr Gemma 4 31B post. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - Long documents carry most of the loss: UD-Q8_K_XL scores KL 0.001 on coding but 0.373 on long documents. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - Tool calling is the second-worst category at every size. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - ik_llama.cpp-only types (IQ5_KS, IQ4_KS) cannot run on mainline llama.cpp, so their frontier spots do not help a stock Metal build. — source: `asserted` - The cache shows only the page text; the Pareto table and per-quant numbers sit in images, so rank order within the frontier is unknown here. — source: `asserted` - Existing Gemma 4 31B result: unsloth holds 8 of 9 frontier spots. Qwen 3.6 27B: bartowski and mradermacher tie at 12 each and unsloth has 10, so no uploader dominates. Both are in the same author's series. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - The benchmark covers 87 Qwen 3.6 27B GGUF quants from 7 uploaders against the unsloth BF16 reference. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - bartowski and mradermacher each hold 12 frontier positions and unsloth holds 10. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - ubergarm's two ik_llama.cpp quants both reach the frontier: IQ5_KS (19.9 GB, KL 0.128) and IQ4_KS (15.8 GB, KL 0.209). — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - Jackrong has no frontier quants out of 11 and lmstudio-community has none out of 3. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - Q8_0 has KL 0.075, close to Qwen 3.6 35B-A3B (0.069) and better than Qwen 3.5 27B (0.120). — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - IQ2_XXS (8.4 GB) keeps 77.5% top-1 agreement with BF16. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark) - The post asks readers to share the URL, not its plots and tables, because the measurements are expensive and subscriptions fund more models. — [source](https://localbench.substack.com/p/qwen-3-6-27b-gguf-quality-benchmark)