Model benchmarks
Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 local LLM performance
As of November 2026, Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 runs at up to 20.5 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
35B
Peak speed
20.5 tok/s
Average speed
20.4 tok/s
Avg PP
331.5 tok/s
Min memory
n/a
Max context
65,536 tokens
Avg output / run
21,701 tokens
Avg runtime / run
18m 50s
Avg quality
83.4
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 | 80.6 | 83.4 | 86.2 |
| Agent Workflow | 81.0 | 86.3 | 91.5 |
| Code Generation | 73.9 | 75.2 | 76.5 |
| Role Play & Narrative | 84.4 | 87.3 | 90.1 |
| Research & Analysis | 83.0 | 84.9 | 86.7 |
Performance by hardware and tool
Every hardware/tool/quantization combination Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 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 GTX 1070 | llama.cpp | — | 20.5 tok/s | 20.4 tok/s | n/a | 65,536 tokens | 83.4 | 2 |
Benchmark runs
All 2 Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 good for coding?
- In our benchmarks, Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 scores 75.2/100 for coding. It runs at about 20.4 tok/s, so if you want more speed, byteshape/Qwen3.8-27B-GGUF:IQ4_XS-ASCII-Condensed is faster (~42.0 tok/s) and still scores well for coding (86.3/100).
- Is Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 good for agentic (tool-using) tasks?
- In our benchmarks, Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 scores 86.3/100 for agentic workflows. It runs at about 20.4 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 Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 for local inference?
- Across 2 community benchmark runs, Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3 reaches up to 20.5 tok/s and averages 20.4 tok/s, with the fastest results on NVIDIA GeForce GTX 1070.
- Which tools have been used to run Huihui-Ornith-1.5-35B-A3B-abliterated-PXQN3?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.