Is Qwen3.8-27B-Q8_0 good for coding?

In our benchmarks, Qwen3.8-27B-Q8_0 scores 72.3/100 for coding. It runs at about 59.2 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-Q8_0 good for agentic (tool-using) tasks?

In our benchmarks, Qwen3.8-27B-Q8_0 scores 77.1/100 for agentic workflows. It runs at about 59.2 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).

How fast is Qwen3.8-27B-Q8_0 for local inference?

Across 24 community benchmark runs, Qwen3.8-27B-Q8_0 reaches up to 77.5 tok/s and averages 59.2 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.

How much memory does Qwen3.8-27B-Q8_0 need?

The leanest observed configuration used about 34.1 GB of memory (quantizations tested: Q8_0).

Which tools have been used to run Qwen3.8-27B-Q8_0?

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

Model benchmarks

Qwen3.8-27B-Q8_0 local LLM performance

As of October 2026, Qwen3.8-27B-Q8_0 runs at up to 77.5 tok/s for local inference (best of 24 community benchmark runs across 2 GPUs).

llama.cppoMLXQ8_0

ShareReddit X

Model size

27B

Peak speed

77.5 tok/s

Average speed

59.2 tok/s

Avg PP

459.2 tok/s

Min memory

34.1 GB

Max context

262,144 tokens

Avg output / run

30,174 tokens

Avg runtime / run

8m 42s

Avg quality

77.6

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 24 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	13.6	77.6	88.7
Agent Workflow	17.9	77.1	92.1
Code Generation	9.4	72.3	86.2
Role Play & Narrative	12.7	82.7	93.0
Research & Analysis	11.6	78.5	90.1

Performance by hardware and tool

Every hardware/tool/quantization combination Qwen3.8-27B-Q8_0 has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
AMD Radeon RX 7900 XTX	llama.cpp	Q8_0	77.5 tok/s	59.6 tok/s	n/a	65,536 tokens	76.8	22
Apple M5 Max	oMLX	Q8_0	54.6 tok/s	53.9 tok/s	34.1 GB	262,144 tokens	86.5	2

Benchmark runs

All 24 Qwen3.8-27B-Q8_0 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-Q8_0 good for coding?: In our benchmarks, Qwen3.8-27B-Q8_0 scores 72.3/100 for coding. It runs at about 59.2 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-Q8_0 good for agentic (tool-using) tasks?: In our benchmarks, Qwen3.8-27B-Q8_0 scores 77.1/100 for agentic workflows. It runs at about 59.2 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).
How fast is Qwen3.8-27B-Q8_0 for local inference?: Across 24 community benchmark runs, Qwen3.8-27B-Q8_0 reaches up to 77.5 tok/s and averages 59.2 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.
How much memory does Qwen3.8-27B-Q8_0 need?: The leanest observed configuration used about 34.1 GB of memory (quantizations tested: Q8_0).
Which tools have been used to run Qwen3.8-27B-Q8_0?: Benchmarks were submitted using llama.cpp, oMLX. Results are community-contributed and updated as new runs arrive.