Model benchmarks
Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX local LLM performance
As of October 2026, Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX runs at up to 37.7 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
27B
Peak speed
37.7 tok/s
Average speed
26.3 tok/s
Avg PP
144.3 tok/s
Min memory
16.3 GB
Max context
131,072 tokens
Avg output / run
61,765 tokens
Avg runtime / run
29m 39s
Avg quality
63.5
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 | 58.6 | 63.5 | 68.3 |
| Agent Workflow | 68.4 | 78.8 | 89.2 |
| Code Generation | 0.9 | 8.5 | 16.2 |
| Role Play & Narrative | 66.5 | 80.3 | 94.1 |
| Research & Analysis | 83.4 | 86.3 | 89.1 |
Performance by hardware and tool
Every hardware/tool/quantization combination Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Apple M4 Pro | oMLX | 4bit | 37.7 tok/s | 26.3 tok/s | 16.3 GB | 131,072 tokens | 63.5 | 2 |
Benchmark runs
All 2 Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX good for coding?
- In our benchmarks, Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX scores 8.5/100 for coding. It runs at about 26.3 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 Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX good for agentic (tool-using) tasks?
- In our benchmarks, Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX scores 78.8/100 for agentic workflows. It runs at about 26.3 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 Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX for local inference?
- Across 2 community benchmark runs, Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX reaches up to 37.7 tok/s and averages 26.3 tok/s, with the fastest results on Apple M4 Pro.
- How much memory does Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX need?
- The leanest observed configuration used about 16.3 GB of memory (quantizations tested: 4bit).
- Which tools have been used to run Qwen3.8-27B-AEON-ULTIMATE-UNCENSORED-4bit-MLX?
- Benchmarks were submitted using oMLX. Results are community-contributed and updated as new runs arrive.