Model benchmarks
Qwen3.8-27b-q4s local LLM performance
As of October 2026, Qwen3.8-27b-q4s runs at up to 67.3 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
27B
Peak speed
67.3 tok/s
Average speed
65.7 tok/s
Avg PP
538.1 tok/s
Min memory
18.6 GB
Max context
200,000 tokens
Avg output / run
16,301 tokens
Avg runtime / run
4m 8s
Avg quality
83.1
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 | 77.2 | 83.1 | 88.9 |
| Agent Workflow | 76.2 | 83.5 | 90.8 |
| Code Generation | 58.4 | 70.0 | 81.6 |
| Role Play & Narrative | 90.3 | 92.6 | 94.9 |
| Research & Analysis | 84.1 | 86.3 | 88.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination Qwen3.8-27b-q4s 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 |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 4090 D | llama.cpp | — | 67.3 tok/s | 65.7 tok/s | 18.6 GB | 200,000 tokens | 83.1 | 2 |
Benchmark runs
All 2 Qwen3.8-27b-q4s runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Qwen3.8-27b-q4s good for coding?
- In our benchmarks, Qwen3.8-27b-q4s scores 70.0/100 for coding. It runs at about 65.7 tok/s, so if you want more speed, Nex-N2.5-mini-IQ3_XXS is faster (~91.4 tok/s) and still scores well for coding (81.6/100).
- Is Qwen3.8-27b-q4s good for agentic (tool-using) tasks?
- In our benchmarks, Qwen3.8-27b-q4s scores 83.5/100 for agentic workflows. It runs at about 65.7 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-q4s for local inference?
- Across 2 community benchmark runs, Qwen3.8-27b-q4s reaches up to 67.3 tok/s and averages 65.7 tok/s, with the fastest results on NVIDIA GeForce RTX 4090 D.
- How much memory does Qwen3.8-27b-q4s need?
- The leanest observed configuration used about 18.6 GB of memory.
- Which tools have been used to run Qwen3.8-27b-q4s?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.