Ornith-1.5-9B
Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M ornith-ai/Ornith-1.5-9B-GGUF | 5.38 GiB | 1 | 6.6 GiB+ | 8 GB GPU | HF files |
| Q5_K_M ornith-ai/Ornith-1.5-9B-GGUF | 6.19 GiB | 1 | 7.4 GiB+ | 8 GB GPU | HF files |
| Q8_0 ornith-ai/Ornith-1.5-9B-GGUF | 9.11 GiB | 1 | 10.4 GiB+ | 12 GB GPU | HF files |
| FP16 mradermacher/Ornith-1.5-9B-GGUF | 16.69 GiB | 1 | 17.9 GiB+ | 20 GB GPU | HF files |
| MLX 4-bit ornith-ai/Ornith-1.5-9B-MLX-4bit | 4.69 GiB | 1 | 5.9 GiB+ | 16 GB Mac | HF files |
| MLX 8-bit ornith-ai/Ornith-1.5-9B-MLX-8bit | 8.86 GiB | 2 | 10.1 GiB+ | 16 GB Mac | HF files |
Sizes are exact file bytes (text weights; projectors listed separately). “+” = recurrent state or other terms unknown. Smallest fit adds 0.5 GiB spare; Macs use 75% of memory.
Quality
Artificial Analysis Intelligence Index
Unknown
Arena text rating
Unknown
Scores describe the evaluated hosted variant, not each quantized file. AA and Arena are never averaged.
Reported speeds 2
| Hardware | Quant | Backend | Prompt t/s | Gen t/s | Date | Source |
|---|---|---|---|---|---|---|
| Steam Deck (AMD Custom GPU 0932 / RADV VANGOGH, Vulkan) | Q4_K_M | llama.cpp Vulkan | 93.9 | 11.2 | 2026-08-20 | llama.cpp #10879 |
| Steam Deck (AMD Custom GPU 0932 / RADV VANGOGH, Vulkan) | Q4_K_M | llama.cpp Vulkan | 98.7 | 11.5 | 2026-08-20 | llama.cpp #10879 |
Reported by the linked sources, not reproduced by LocalFit. All rows in the speed database
Community threads
- The reasoning speed of this model is really fast, but the quality is just average. Hugging Face · 2026-09-16
- Model thinks very, VERY long Hugging Face · 2026-10-11
- https://www.reddit.com/r/LocalLLaMA/comments/1wha6sv/i_benchmarked_ifmk2horizon7b_on_16gb_vram/ Reddit · 2026-09-15
Qualitative signals only; titles are not verified claims.