Model benchmarks

Qwen3.6:35B-4bit local LLM performance

As of July 2026, Qwen3.6:35B-4bit runs at up to 103.4 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).

oMLX
ShareRedditX

Model size

35B

Peak speed

103.4 tok/s

Average speed

64.6 tok/s

Min memory

21.5 GB

Max context

65,536 tokens

Avg quality

80.9

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.

TaskP5 (low)AvgP95 (high)
Overall79.680.982.1
Agent Workflow80.783.385.8
Code Generation74.577.380.1
Role Play & Narrative75.780.986.1
Research & Analysis81.582.082.5

Performance by hardware and tool

Every hardware/tool/quantization combination Qwen3.6:35B-4bit has been benchmarked on, ranked by peak token generation speed. Last updated July 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
Apple M5 MaxoMLX103.4 tok/s64.6 tok/s21.5 GB65,536 tokens80.92

Frequently asked questions

Is Qwen3.6:35B-4bit good for coding?
In our benchmarks, Qwen3.6:35B-4bit scores 77.3/100 for coding. It runs at about 64.6 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 Qwen3.6:35B-4bit good for agentic (tool-using) tasks?
In our benchmarks, Qwen3.6:35B-4bit scores 83.3/100 for agentic workflows. It runs at about 64.6 tok/s, so if you want more speed, unsloth/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF:UD-Q4_K_XL is faster (~169.2 tok/s) and still scores well for agentic workflows (83.0/100).
How fast is Qwen3.6:35B-4bit for local inference?
Across 2 community benchmark runs, Qwen3.6:35B-4bit reaches up to 103.4 tok/s and averages 64.6 tok/s, with the fastest results on Apple M5 Max.
How much memory does Qwen3.6:35B-4bit need?
The leanest observed configuration used about 21.5 GB of memory.
Which tools have been used to run Qwen3.6:35B-4bit?
Benchmarks were submitted using oMLX. Results are community-contributed and updated as new runs arrive.