Model benchmarks
qwen36-UD-Q4_K_XL local LLM performance
As of November 2026, qwen36-UD-Q4_K_XL runs at up to 58.6 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
58.6 tok/s
Average speed
57.6 tok/s
Avg PP
599.9 tok/s
Min memory
25.5 GB
Max context
65,536 tokens
Avg output / run
18,728 tokens
Avg runtime / run
5m 41s
Avg quality
78.9
Benchmark runs
2
GPUs tested
1
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 2 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 | 74.7 | 78.9 | 83.0 |
| Agent Workflow | 74.9 | 78.6 | 82.3 |
| Code Generation | 68.8 | 72.2 | 75.5 |
| Role Play & Narrative | 83.2 | 87.7 | 92.2 |
| Research & Analysis | 71.9 | 76.9 | 81.9 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen36-UD-Q4_K_XL 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 RX 7900 XT | Strata | Q4_K | 58.6 tok/s | 58.6 tok/s | n/a | 65,536 tokens | 83.4 | 1 |
| AMD Radeon RX 7900 XT | Unsloth | Q4_K | 56.6 tok/s | 56.6 tok/s | 25.5 GB | 65,536 tokens | 74.3 | 1 |
Benchmark runs
All 2 qwen36-UD-Q4_K_XL runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen36-ud-q4_k_xl good for coding?
- In our benchmarks, qwen36-ud-q4_k_xl scores 72.2/100 for coding. It runs at about 57.6 tok/s, so if you want more speed, swift-1.5-27b is faster (~96.8 tok/s) and still scores well for coding (84.8/100).
- Is qwen36-ud-q4_k_xl good for agentic (tool-using) tasks?
- In our benchmarks, qwen36-ud-q4_k_xl scores 78.6/100 for agentic workflows. It runs at about 57.6 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 qwen36-UD-Q4_K_XL for local inference?
- Across 2 community benchmark runs, qwen36-UD-Q4_K_XL reaches up to 58.6 tok/s and averages 57.6 tok/s, with the fastest results on AMD Radeon RX 7900 XT.
- How much memory does qwen36-UD-Q4_K_XL need?
- The leanest observed configuration used about 25.5 GB of memory (quantizations tested: Q4_K).
- Which tools have been used to run qwen36-UD-Q4_K_XL?
- Benchmarks were submitted using Strata, Unsloth. Results are community-contributed and updated as new runs arrive.