Model benchmarks
qwen3.6:27b-coding local LLM performance
As of November 2026, qwen3.6:27b-coding runs at up to 24.3 tok/s for local inference (best of 3 community benchmark runs across 2 GPUs).
Model size
27.3B
Peak speed
24.3 tok/s
Average speed
20.2 tok/s
Avg PP
326.7 tok/s
Min memory
16.8 GB
Max context
16,384 tokens
Avg output / run
16,575 tokens
Avg runtime / run
15m 1s
Avg quality
81.0
Benchmark runs
3
GPUs tested
2
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 3 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 | 78.9 | 81.0 | 84.0 |
| Agent Workflow | 82.9 | 87.0 | 90.3 |
| Code Generation | 70.6 | 72.6 | 74.6 |
| Role Play & Narrative | 75.0 | 80.8 | 88.3 |
| Research & Analysis | 82.0 | 83.6 | 85.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen3.6:27b-coding has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| AMD Radeon 8060S | Ollama | Q4_K_M | 24.3 tok/s | 23.5 tok/s | n/a | 16,384 tokens | 81.6 | 2 |
| AMD Radeon RX 9070/9070 XT/9070 GRE | Ollama | Q4_K_M | 13.7 tok/s | 13.7 tok/s | 16.8 GB | 8,192 tokens | 79.7 | 1 |
Benchmark runs
All 3 qwen3.6:27b-coding runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen3.6:27b-coding good for coding?
- In our benchmarks, qwen3.6:27b-coding scores 72.6/100 for coding. It runs at about 20.2 tok/s, so if you want more speed, Huihui-Qwen3.8-27B-abliterated-Swift-1.5-GSQ-RCO-IQ3_S-mtp is faster (~43.5 tok/s) and still scores well for coding (86.6/100).
- Is qwen3.6:27b-coding good for agentic (tool-using) tasks?
- In our benchmarks, qwen3.6:27b-coding scores 87.0/100 for agentic workflows. It runs at about 20.2 tok/s, so if you want more speed, muse-glimmer:latest is faster (~33.3 tok/s) and still scores well for agentic workflows (91.6/100).
- How fast is qwen3.6:27b-coding for local inference?
- Across 3 community benchmark runs, qwen3.6:27b-coding reaches up to 24.3 tok/s and averages 20.2 tok/s, with the fastest results on AMD Radeon 8060S.
- How much memory does qwen3.6:27b-coding need?
- The leanest observed configuration used about 16.8 GB of memory (quantizations tested: Q4_K_M).
- Which tools have been used to run qwen3.6:27b-coding?
- Benchmarks were submitted using Ollama. Results are community-contributed and updated as new runs arrive.