Model benchmarks
ornith-ai/Ornith-1.5-35B-A3B-GGUF local LLM performance
As of November 2026, ornith-ai/Ornith-1.5-35B-A3B-GGUF runs at up to 98.7 tok/s for local inference (best of 5 community benchmark runs across 2 GPUs).
Model size
35B
Peak speed
98.7 tok/s
Average speed
80.3 tok/s
Avg PP
900.7 tok/s
Min memory
19.9 GB
Max context
262,144 tokens
Avg output / run
40,955 tokens
Avg runtime / run
9m 4s
Avg quality
83.3
Benchmark runs
5
GPUs tested
2
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 5 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 | 74.9 | 83.3 | 87.3 |
| Agent Workflow | 41.2 | 75.6 | 90.8 |
| Code Generation | 76.5 | 82.1 | 85.6 |
| Role Play & Narrative | 86.3 | 89.8 | 93.1 |
| Research & Analysis | 83.0 | 85.4 | 88.1 |
Performance by hardware and tool
Every hardware/tool/quantization combination ornith-ai/Ornith-1.5-35B-A3B-GGUF 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 |
|---|---|---|---|---|---|---|---|---|
| AMD Radeon RX 7900 XTX | Unsloth Studio | — | 98.7 tok/s | 96.3 tok/s | n/a | 131,072 tokens | 80.1 | 2 |
| AMD Radeon RX 7900 XTX | Unsloth Studio | Q8_0 | 89.9 tok/s | 89.9 tok/s | n/a | 131,072 tokens | 86.0 | 1 |
| AMD Radeon RX 7900 XTX | Unsloth Studio | Q6_K | 80.7 tok/s | 80.7 tok/s | n/a | 262,144 tokens | 86.3 | 1 |
| NVIDIA GeForce RTX 5050 Laptop GPU | Unsloth Studio | IQ2_M | 38.2 tok/s | 38.2 tok/s | 19.9 GB | 32,768 tokens | 83.7 | 1 |
Benchmark runs
All 5 ornith-ai/Ornith-1.5-35B-A3B-GGUF runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is ornith-ai/Ornith-1.5-35B-A3B-GGUF good for coding?
- In our benchmarks, ornith-ai/Ornith-1.5-35B-A3B-GGUF scores 82.1/100 for coding. It runs at about 80.3 tok/s, so if you want more speed, Ornith-1.5-35B-A3B.APEX-I-NanoPlus is faster (~96.1 tok/s) and still scores well for coding (81.8/100).
- Is ornith-ai/Ornith-1.5-35B-A3B-GGUF good for agentic (tool-using) tasks?
- In our benchmarks, ornith-ai/Ornith-1.5-35B-A3B-GGUF scores 75.6/100 for agentic workflows. It runs at about 80.3 tok/s, so if you want more speed, qwen3.8-flash-next-iq3_s is faster (~99.8 tok/s) and still scores well for agentic workflows (90.2/100).
- How fast is ornith-ai/Ornith-1.5-35B-A3B-GGUF for local inference?
- Across 5 community benchmark runs, ornith-ai/Ornith-1.5-35B-A3B-GGUF reaches up to 98.7 tok/s and averages 80.3 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.
- How much memory does ornith-ai/Ornith-1.5-35B-A3B-GGUF need?
- The leanest observed configuration used about 19.9 GB of memory (quantizations tested: IQ2_M, Q6_K, Q8_0).
- Which tools have been used to run ornith-ai/Ornith-1.5-35B-A3B-GGUF?
- Benchmarks were submitted using Unsloth Studio. Results are community-contributed and updated as new runs arrive.