Model benchmarks
qwen3.5-4B-super-coder.Q4_0 local LLM performance
As of November 2026, qwen3.5-4B-super-coder.Q4_0 runs at up to 66.4 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
4B
Peak speed
66.4 tok/s
Average speed
66.0 tok/s
Avg PP
1062.9 tok/s
Min memory
3.9 GB
Max context
32,768 tokens
Avg output / run
6,134 tokens
Avg runtime / run
1m 37s
Avg quality
50.3
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 | 46.1 | 50.3 | 53.5 |
| Agent Workflow | 16.6 | 46.4 | 64.3 |
| Code Generation | 19.8 | 30.0 | 38.9 |
| Role Play & Narrative | 56.6 | 69.8 | 80.9 |
| Research & Analysis | 48.6 | 55.1 | 64.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen3.5-4B-super-coder.Q4_0 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 GeForce RTX 4060 Laptop GPU | llama.cpp | Q4_0 | 66.4 tok/s | 66.0 tok/s | 3.9 GB | 32,768 tokens | 50.3 | 3 |
Benchmark runs
All 3 qwen3.5-4B-super-coder.Q4_0 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen3.5-4B-super-coder.Q4_0 good for coding?
- In our benchmarks, qwen3.5-4B-super-coder.Q4_0 scores 30.0/100 for coding. It runs at about 66.0 tok/s, so if you want more speed, Ornith-1.5-35B-A3B.APEX-I-NanoPlus is faster (~110.5 tok/s) and still scores well for coding (81.1/100).
- Is qwen3.5-4B-super-coder.Q4_0 good for agentic (tool-using) tasks?
- In our benchmarks, qwen3.5-4B-super-coder.Q4_0 scores 46.4/100 for agentic workflows. It runs at about 66.0 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 qwen3.5-4B-super-coder.Q4_0 for local inference?
- Across 3 community benchmark runs, qwen3.5-4B-super-coder.Q4_0 reaches up to 66.4 tok/s and averages 66.0 tok/s, with the fastest results on NVIDIA GeForce RTX 4060 Laptop GPU.
- How much memory does qwen3.5-4B-super-coder.Q4_0 need?
- The leanest observed configuration used about 3.9 GB of memory (quantizations tested: Q4_0).
- Which tools have been used to run qwen3.5-4B-super-coder.Q4_0?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.