Model benchmarks
qwen3.8-large-ctx:latest local LLM performance
As of November 2026, qwen3.8-large-ctx:latest runs at up to 25.9 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
27.3B
Peak speed
25.9 tok/s
Average speed
25.6 tok/s
Avg PP
205.5 tok/s
Min memory
14.0 GB
Max context
81,920 tokens
Avg output / run
14,308 tokens
Avg runtime / run
9m 41s
Avg quality
82.8
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 | 81.8 | 82.8 | 83.9 |
| Agent Workflow | 76.6 | 82.7 | 88.7 |
| Code Generation | 78.2 | 83.3 | 88.4 |
| Role Play & Narrative | 78.2 | 80.6 | 82.9 |
| Research & Analysis | 84.0 | 84.8 | 85.6 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen3.8-large-ctx:latest 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 7600 XT | Ollama | IQ3_S | 25.9 tok/s | 25.6 tok/s | 14.0 GB | 81,920 tokens | 82.8 | 2 |
Benchmark runs
All 2 qwen3.8-large-ctx:latest runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen3.8-large-ctx:latest good for coding?
- In our benchmarks, qwen3.8-large-ctx:latest scores 83.3/100 for coding, among the top 3 for coding on consumer hardware (≤24 GB VRAM). It runs at about 25.6 tok/s, so if you want more speed, Qwen3.8-27B-Turbo-Brilliance-50X-RI-GSQ-RCO-IQ3_XXS-mtp is faster (~37.5 tok/s) and still scores well for coding (83.8/100).
- Is qwen3.8-large-ctx:latest good for agentic (tool-using) tasks?
- In our benchmarks, qwen3.8-large-ctx:latest scores 82.7/100 for agentic workflows. It runs at about 25.6 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 qwen3.8-large-ctx:latest for local inference?
- Across 2 community benchmark runs, qwen3.8-large-ctx:latest reaches up to 25.9 tok/s and averages 25.6 tok/s, with the fastest results on AMD Radeon RX 7600 XT.
- How much memory does qwen3.8-large-ctx:latest need?
- The leanest observed configuration used about 14.0 GB of memory (quantizations tested: IQ3_S).
- Which tools have been used to run qwen3.8-large-ctx:latest?
- Benchmarks were submitted using Ollama. Results are community-contributed and updated as new runs arrive.