Model benchmarks
mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF local LLM performance
As of August 2026, mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF runs at up to 193.7 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
vLLM
Model size
35B
Peak speed
193.7 tok/s
Average speed
190.9 tok/s
Min memory
24.2 GB
Max context
21,235 tokens
Best quality
89.3
Benchmark runs
2
GPUs tested
1
Performance by hardware and tool
Every hardware/tool/quantization combination mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 5090 | vLLM | — | 193.7 tok/s | 190.9 tok/s | 24.2 GB | 21,235 tokens | 82.1 | 2 |
Frequently asked questions
- How fast is mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF for local inference?
- Across 2 community benchmark runs, mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF reaches up to 193.7 tok/s and averages 190.9 tok/s, with the fastest results on NVIDIA GeForce RTX 5090.
- How much memory does mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF need?
- The leanest observed configuration used about 24.2 GB of memory.
- Which tools have been used to run mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF?
- Benchmarks were submitted using vLLM. Results are community-contributed and updated as new runs arrive.