Model benchmarks
qwen3.5-4b@q4_k_m local LLM performance
As of November 2026, qwen3.5-4b@q4_k_m runs at up to 52.2 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
4B
Peak speed
52.2 tok/s
Average speed
50.6 tok/s
Avg PP
546.1 tok/s
Min memory
4.0 GB
Max context
65,536 tokens
Avg output / run
17,503 tokens
Avg runtime / run
6m 12s
Avg quality
67.4
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 | 62.5 | 67.4 | 72.3 |
| Agent Workflow | 66.1 | 72.7 | 79.2 |
| Code Generation | 57.9 | 58.1 | 58.2 |
| Role Play & Narrative | 73.9 | 78.6 | 83.4 |
| Research & Analysis | 42.5 | 60.4 | 78.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen3.5-4b@q4_k_m 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 |
|---|---|---|---|---|---|---|---|---|
| Apple M1 Max | LM Studio | Q4_K_M | 52.2 tok/s | 50.6 tok/s | 4.0 GB | 65,536 tokens | 67.4 | 2 |
Benchmark runs
All 2 qwen3.5-4b@q4_k_m runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen3.5-4b@q4_k_m good for coding?
- In our benchmarks, qwen3.5-4b@q4_k_m scores 58.1/100 for coding. It runs at about 50.6 tok/s, so if you want more speed, nvidia/Qwen3.8-27B-NVFP4 is faster (~63.3 tok/s) and still scores well for coding (82.7/100).
- Is qwen3.5-4b@q4_k_m good for agentic (tool-using) tasks?
- In our benchmarks, qwen3.5-4b@q4_k_m scores 72.7/100 for agentic workflows. It runs at about 50.6 tok/s, so if you want more speed, swift-iq2_xs is faster (~91.5 tok/s) and still scores well for agentic workflows (91.4/100).
- How fast is qwen3.5-4b@q4_k_m for local inference?
- Across 2 community benchmark runs, qwen3.5-4b@q4_k_m reaches up to 52.2 tok/s and averages 50.6 tok/s, with the fastest results on Apple M1 Max.
- How much memory does qwen3.5-4b@q4_k_m need?
- The leanest observed configuration used about 4.0 GB of memory (quantizations tested: Q4_K_M).
- Which tools have been used to run qwen3.5-4b@q4_k_m?
- Benchmarks were submitted using LM Studio. Results are community-contributed and updated as new runs arrive.