Model benchmarks
Hemmingway-1-oQ8e-mtp local LLM performance
As of September 2026, Hemmingway-1-oQ8e-mtp runs at up to 32.0 tok/s for local inference (best of 4 community benchmark runs across 1 GPU).
Model size
27B
Peak speed
32.0 tok/s
Average speed
28.2 tok/s
Avg PP
397.3 tok/s
Min memory
28.4 GB
Max context
262,144 tokens
Avg output / run
51,683 tokens
Avg runtime / run
31m 59s
Avg quality
85.4
Benchmark runs
4
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 4 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 | 82.5 | 85.4 | 87.5 |
| Agent Workflow | 78.3 | 85.0 | 90.8 |
| Code Generation | 74.9 | 76.0 | 77.5 |
| Role Play & Narrative | 90.9 | 93.3 | 94.6 |
| Research & Analysis | 84.6 | 87.3 | 90.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination Hemmingway-1-oQ8e-mtp has been benchmarked on, ranked by peak token generation speed. Last updated September 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Apple M5 Max | oMLX | oQ8e | 32.0 tok/s | 28.2 tok/s | 28.4 GB | 262,144 tokens | 85.4 | 4 |
Benchmark runs
All 4 Hemmingway-1-oQ8e-mtp runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Hemmingway-1-oQ8e-mtp good for coding?
- In our benchmarks, Hemmingway-1-oQ8e-mtp scores 76.0/100 for coding. It runs at about 28.2 tok/s, so if you want more speed, local-qwen is faster (~60.2 tok/s) and still scores well for coding (84.8/100).
- Is Hemmingway-1-oQ8e-mtp good for agentic (tool-using) tasks?
- In our benchmarks, Hemmingway-1-oQ8e-mtp scores 85.0/100 for agentic workflows. It runs at about 28.2 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).
- How fast is Hemmingway-1-oQ8e-mtp for local inference?
- Across 4 community benchmark runs, Hemmingway-1-oQ8e-mtp reaches up to 32.0 tok/s and averages 28.2 tok/s, with the fastest results on Apple M5 Max.
- How much memory does Hemmingway-1-oQ8e-mtp need?
- The leanest observed configuration used about 28.4 GB of memory (quantizations tested: oQ8e).
- Which tools have been used to run Hemmingway-1-oQ8e-mtp?
- Benchmarks were submitted using oMLX. Results are community-contributed and updated as new runs arrive.