Model Anatomy See what's inside open-weight AI models

Can I run Holo4-27B on an RTX 5090?

Yes, with a GGUF version

The published weights are too big, but the Q4_K_M file from Hcompany/Holo4-27B-GGUF (16.9 GB) fits, with room for ≈ 175K tokens of context.

Full analysis of Holo4-27B Everything that runs on RTX 5090 Get the Q4_K_M file on Hugging Face
ollama run hf.co/Hcompany/Holo4-27B-GGUF:Q4_K_M

Files and context on RTX 5090

Each file of Holo4-27B (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

FileSizeOn RTX 5090 Context (16-bit)Context (8-bit)
As published54.7 GBDoesn't fit——
Q4_K_MHcompany/Holo4-27B-GGUF16.9 GBFits175K256K

Holo4-27B on other hardware

Popular models that run on RTX 5090

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.