Model benchmarks
a3-qwen3.5-9b local LLM performance
As of November 2026, a3-qwen3.5-9b runs at up to 119.4 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
9B
Peak speed
119.4 tok/s
Average speed
114.8 tok/s
Avg PP
2939.0 tok/s
Min memory
n/a
Max context
65,536 tokens
Avg output / run
10,815 tokens
Avg runtime / run
1m 31s
Avg quality
66.2
Benchmark runs
3
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 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 | 62.0 | 66.2 | 72.5 |
| Agent Workflow | 53.0 | 65.0 | 75.2 |
| Code Generation | 55.2 | 61.1 | 65.1 |
| Role Play & Narrative | 59.0 | 67.1 | 73.9 |
| Research & Analysis | 62.8 | 71.5 | 79.0 |
Performance by hardware and tool
Every hardware/tool/quantization combination a3-qwen3.5-9b 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 |
|---|---|---|---|---|---|---|---|---|
| NVIDIA A10G | ninfer | — | 119.4 tok/s | 114.8 tok/s | n/a | 65,536 tokens | 66.2 | 3 |
Benchmark runs
All 3 a3-qwen3.5-9b runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is a3-qwen3.5-9b good for coding?
- In our benchmarks, a3-qwen3.5-9b scores 61.1/100 for coding. It runs at about 114.8 tok/s, so if you want more speed, Qwen3.6-35B-A3B-Q4_K_XL is faster (~156.9 tok/s) and still scores well for coding (82.2/100).
- Is a3-qwen3.5-9b good for agentic (tool-using) tasks?
- In our benchmarks, a3-qwen3.5-9b scores 65.0/100 for agentic workflows. It runs at about 114.8 tok/s, so if you want more speed, swift-1.5-qwen3.8-flash-next-iq3_s is faster (~259.0 tok/s) and still scores well for agentic workflows (88.9/100).
- How fast is a3-qwen3.5-9b for local inference?
- Across 3 community benchmark runs, a3-qwen3.5-9b reaches up to 119.4 tok/s and averages 114.8 tok/s, with the fastest results on NVIDIA A10G.
- Which tools have been used to run a3-qwen3.5-9b?
- Benchmarks were submitted using ninfer. Results are community-contributed and updated as new runs arrive.