Is Qwen3.8-27B-Q4_K_XL good for coding?

In our benchmarks, Qwen3.8-27B-Q4_K_XL scores 78.7/100 for coding. It runs at about 72.4 tok/s, so if you want more speed, ornith-ai/Ornith-1.5-35B-A3B-GGUF is faster (~90.8 tok/s) and still scores well for coding (83.8/100).

Is Qwen3.8-27B-Q4_K_XL good for agentic (tool-using) tasks?

In our benchmarks, Qwen3.8-27B-Q4_K_XL scores 76.5/100 for agentic workflows. It runs at about 72.4 tok/s, so if you want more speed, Nex-N2.5-mini-APEX-Mini is faster (~88.2 tok/s) and still scores well for agentic workflows (91.3/100).

How fast is Qwen3.8-27B-Q4_K_XL for local inference?

Across 15 community benchmark runs, Qwen3.8-27B-Q4_K_XL reaches up to 77.0 tok/s and averages 72.4 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.

Which tools have been used to run Qwen3.8-27B-Q4_K_XL?

Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.

Model benchmarks

Qwen3.8-27B-Q4_K_XL local LLM performance

As of September 2026, Qwen3.8-27B-Q4_K_XL runs at up to 77.0 tok/s for local inference (best of 15 community benchmark runs across 1 GPU).

llama.cppQ4_K

ShareReddit X

Model size

27B

Peak speed

77.0 tok/s

Average speed

72.4 tok/s

Min memory

n/a

Max context

65,536 tokens

Avg output / run

15,386 tokens

Avg runtime / run

3m 30s

Avg quality

81.5

Benchmark runs

GPUs tested

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 15 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

Task	P5 (low)	Avg	P95 (high)
Overall	72.3	81.5	86.7
Agent Workflow	51.4	76.5	89.6
Code Generation	68.6	78.7	86.5
Role Play & Narrative	75.4	85.7	92.3
Research & Analysis	78.4	85.0	89.3

Performance by hardware and tool

Every hardware/tool/quantization combination Qwen3.8-27B-Q4_K_XL has been benchmarked on, ranked by peak token generation speed. Last updated September 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
AMD Radeon RX 7900 XTX	llama.cpp	Q4_K	77.0 tok/s	72.4 tok/s	n/a	65,536 tokens	81.5	15

Benchmark runs

All 15 Qwen3.8-27B-Q4_K_XL runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Open run page →

Frequently asked questions

Is Qwen3.8-27B-Q4_K_XL good for coding?: In our benchmarks, Qwen3.8-27B-Q4_K_XL scores 78.7/100 for coding. It runs at about 72.4 tok/s, so if you want more speed, ornith-ai/Ornith-1.5-35B-A3B-GGUF is faster (~90.8 tok/s) and still scores well for coding (83.8/100).
Is Qwen3.8-27B-Q4_K_XL good for agentic (tool-using) tasks?: In our benchmarks, Qwen3.8-27B-Q4_K_XL scores 76.5/100 for agentic workflows. It runs at about 72.4 tok/s, so if you want more speed, Nex-N2.5-mini-APEX-Mini is faster (~88.2 tok/s) and still scores well for agentic workflows (91.3/100).
How fast is Qwen3.8-27B-Q4_K_XL for local inference?: Across 15 community benchmark runs, Qwen3.8-27B-Q4_K_XL reaches up to 77.0 tok/s and averages 72.4 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.
Which tools have been used to run Qwen3.8-27B-Q4_K_XL?: Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.