Model benchmarks
Ornith-1.5-9B-IQ3_M-GGUF local LLM performance
As of October 2026, Ornith-1.5-9B-IQ3_M-GGUF runs at up to 104.4 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
9B
Peak speed
104.4 tok/s
Average speed
91.0 tok/s
Avg PP
2400.2 tok/s
Min memory
11.6 GB
Max context
262,144 tokens
Avg output / run
26,724 tokens
Avg runtime / run
5m 58s
Avg quality
75.2
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 | 69.2 | 75.2 | 81.3 |
| Agent Workflow | 72.2 | 73.5 | 74.8 |
| Code Generation | 51.5 | 65.1 | 78.7 |
| Role Play & Narrative | 80.5 | 84.4 | 88.3 |
| Research & Analysis | 72.6 | 78.0 | 83.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination Ornith-1.5-9B-IQ3_M-GGUF 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 |
|---|---|---|---|---|---|---|---|---|
| NVIDIA RTX A6000 | llama.cpp | — | 104.4 tok/s | 91.0 tok/s | 11.6 GB | 262,144 tokens | 75.2 | 2 |
Benchmark runs
All 2 Ornith-1.5-9B-IQ3_M-GGUF runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Ornith-1.5-9B-IQ3_M-GGUF good for coding?
- In our benchmarks, Ornith-1.5-9B-IQ3_M-GGUF scores 65.1/100 for coding. It runs at about 91.0 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 Ornith-1.5-9B-IQ3_M-GGUF good for agentic (tool-using) tasks?
- In our benchmarks, Ornith-1.5-9B-IQ3_M-GGUF scores 73.5/100 for agentic workflows. It runs at about 91.0 tok/s, so if you want more speed, cyber-tiel-coder-35b-a3b-mtp is faster (~115.9 tok/s) and still scores well for agentic workflows (86.4/100).
- How fast is Ornith-1.5-9B-IQ3_M-GGUF for local inference?
- Across 2 community benchmark runs, Ornith-1.5-9B-IQ3_M-GGUF reaches up to 104.4 tok/s and averages 91.0 tok/s, with the fastest results on NVIDIA RTX A6000.
- How much memory does Ornith-1.5-9B-IQ3_M-GGUF need?
- The leanest observed configuration used about 11.6 GB of memory.
- Which tools have been used to run Ornith-1.5-9B-IQ3_M-GGUF?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.