Ornith 1.5 122B VRAM requirements: which GPUs can run it locally?
The smallest GPU memory that runs Ornith 1.5 122B well is 72 GB (Q4_K_M, e.g. RTX PRO 5000 Blackwell 72GB). 15 of 132 GPUs run it well. It's a mixture-of-experts model, so all weights must fit in memory but speed depends only on the active experts.
SWE-bench Verified 84.5
VRAM needed per quantization
| Quantization | Weights | 16K context | 32K context | Full 256K |
|---|---|---|---|---|
| Q4_K_M | 70.6 GB | 71.0 GB | 71.4 GB | 76.5 GB |
| Q8_0 | 123.5 GB | 123.8 GB | 124.2 GB | 129.3 GB |
Weights plus fp16 KV cache. Leave ~1 GB headroom for the inference engine.
Run it locally
Each command downloads the GGUF from Hugging Face, then Ollama starts a chat and llama.cpp starts an OpenAI-compatible server with a web UI on port 8080.
ollama run hf.co/ornith-ai/Ornith-1.5-122B-GGUF:Q4_K_M llama-server -hf ornith-ai/Ornith-1.5-122B-GGUF:Q4_K_M ollama run hf.co/ornith-ai/Ornith-1.5-122B-GGUF:Q8_0 llama-server -hf ornith-ai/Ornith-1.5-122B-GGUF:Q8_0 Cheapest GPUs that can run Ornith 1.5 122B
Lowest launch MSRP among GPUs that run its smallest quantization (Q4_K_M) well, with at least 16K context.
As an Amazon Associate, I earn from qualifying purchases.
GPUs that can run Ornith 1.5 122B
Best quantization that fits on each GPU, with the context and speed it gets. For chips with several memory sizes, the smallest that works is shown.
NVIDIA
- RTX PRO 5000 Blackwell 72GB 72 GB Q4_K_M 59K context · ~129 tok/s
- RTX PRO 5500 Blackwell 84 GB Q4_K_M 256K context · ~132 tok/s
- RTX PRO 6000 Blackwell 96 GB Q4_K_M 256K context · ~155 tok/s
- DGX Spark (GB 10) 128 GB Q8_0 197K context · ~22 tok/s
- RTX Spark 128 GB Q8_0 197K context · ~24 tok/s
AMD
- Ryzen AI Max (Strix Halo) 96 GB Q4_K_M 256K context · ~33 tok/s
- Ryzen AI Max 400 (Gorgon Halo) 128 GB Q8_0 197K context · ~22 tok/s
Apple
- M2 Max 96 GB Q4_K_M 256K context · ~50 tok/s
- M3 Max 96 GB Q4_K_M 256K context · ~38 tok/s
- M3 Ultra 96 GB Q4_K_M 256K context · ~67 tok/s
- M5 Ultra 96 GB Q4_K_M 256K context · ~92 tok/s
- M1 Ultra 128 GB Q8_0 197K context · ~43 tok/s
- M2 Ultra 128 GB Q8_0 197K context · ~43 tok/s
- M4 Max 128 GB Q8_0 197K context · ~42 tok/s
- M5 Max 128 GB Q8_0 197K context · ~47 tok/s