Model benchmarks
Qwen3.8-27B-GSQ-RCO_t1p95k20m0 local LLM performance
As of November 2026, Qwen3.8-27B-GSQ-RCO_t1p95k20m0 runs at up to 41.6 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
27B
Peak speed
41.6 tok/s
Average speed
41.4 tok/s
Avg PP
565.1 tok/s
Min memory
44.3 GB
Max context
32,768 tokens
Avg output / run
44,283 tokens
Avg runtime / run
17m 36s
Avg quality
85.8
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 | 84.2 | 85.8 | 87.4 |
| Agent Workflow | 72.9 | 79.6 | 86.4 |
| Code Generation | 81.6 | 81.7 | 81.8 |
| Role Play & Narrative | 93.5 | 94.1 | 94.7 |
| Research & Analysis | 87.6 | 87.9 | 88.3 |
Performance by hardware and tool
Every hardware/tool/quantization combination Qwen3.8-27B-GSQ-RCO_t1p95k20m0 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 |
|---|---|---|---|---|---|---|---|---|
| Tesla V100-PCIE-32GB | llama.cpp | — | 41.6 tok/s | 41.4 tok/s | 44.3 GB | 32,768 tokens | 85.8 | 2 |
Benchmark runs
All 2 Qwen3.8-27B-GSQ-RCO_t1p95k20m0 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Qwen3.8-27B-GSQ-RCO_t1p95k20m0 good for coding?
- In our benchmarks, Qwen3.8-27B-GSQ-RCO_t1p95k20m0 scores 81.7/100 for coding. It runs at about 41.4 tok/s, so if you want more speed, local-qwen is faster (~60.2 tok/s) and still scores well for coding (84.8/100).
- Is Qwen3.8-27B-GSQ-RCO_t1p95k20m0 good for agentic (tool-using) tasks?
- In our benchmarks, Qwen3.8-27B-GSQ-RCO_t1p95k20m0 scores 79.6/100 for agentic workflows. It runs at about 41.4 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-GSQ-RCO_t1p95k20m0 for local inference?
- Across 2 community benchmark runs, Qwen3.8-27B-GSQ-RCO_t1p95k20m0 reaches up to 41.6 tok/s and averages 41.4 tok/s, with the fastest results on Tesla V100-PCIE-32GB.
- How much memory does Qwen3.8-27B-GSQ-RCO_t1p95k20m0 need?
- The leanest observed configuration used about 44.3 GB of memory.
- Which tools have been used to run Qwen3.8-27B-GSQ-RCO_t1p95k20m0?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.