Ornith-1.5-9B GGUF size and VRAM requirements
ornith-ai/Ornith-1.5-9B-GGUF is a large language model with 9.65 billion parameters, built on the qwen35 architecture. It is released under the mit license and has been downloaded 1,144,037 times.
Under the hood it uses 32 transformer layers, a hidden size of 4,096, 16 attention heads. It uses a hybrid attention design: only 8 of its 32 layers keep a full KV cache, while the rest use linear attention with a constant-size state. Long contexts therefore cost far less memory than on a conventional model of this size.
To run ornith-ai/Ornith-1.5-9B-GGUF locally at a 4,096-token context, its quantized versions need between 6.31 GB (Q4_K_M, lowest quality) and 18.07 GB (BF16, highest quality) of memory, weights plus KV cache and a system margin included. Context length drives that memory directly: at 4,096 tokens the KV cache for Ornith-1.5-9B-GGUF is about 0.12 GB, rising to roughly 8.0 GB at its full 262,144-token context. Shorter prompts free up memory for a higher-quality quantization.
For most users the best balance is Q6_K, needing about 7.96 GB. That means ornith-ai/Ornith-1.5-9B-GGUF fits entirely in the VRAM of an 8 GB GPU or larger, running fully on the GPU.
Available GGUF quantizations for ornith-ai/Ornith-1.5-9B-GGUF include Q4_K_M, Q5_K_M, Q6_K, Q8_0, BF16. The model supports a native context length of up to 262,144 tokens; a longer context grows the KV cache and the memory needed.
GGUF file size and memory by quantization
Compare real GGUF weight sizes, estimated KV cache and total memory for Q4, Q5, Q8 and every quantization published in this repository.
| Quant. | Bits | Quality | Weights | KV | Total | Speed~ | Verdict |
|---|---|---|---|---|---|---|---|
| Q4_K_M | 4.79 | Good | 5.38 GB | 0.12 GB | 6.31 GB | 74.3 t/s | Fits in VRAM |
| Q5_K_M | 5.51 | Very good | 6.19 GB | 0.12 GB | 7.11 GB | 64.7 t/s | Fits in VRAM |
| Q6_K | 6.26 | Very good | 7.04 GB | 0.12 GB | 7.96 GB | 56.8 t/s | Fits in VRAM |
| Q8_0 | 8.11 | Excellent | 9.11 GB | 0.12 GB | 10.04 GB | 5.5 t/s | Offload |
| BF16 | 15.26 | Excellent | 17.14 GB | 0.12 GB | 18.07 GB | 2.9 t/s | Offload |
KV cache computed from the model's exact architecture. Speed is a rough estimate bounded by memory bandwidth.
Frequently asked questions
What kind of model is ornith-ai/Ornith-1.5-9B-GGUF?
ornith-ai/Ornith-1.5-9B-GGUF is a language model with 9.65 billion parameters, based on the qwen35 architecture. It is released under the mit license and distributed as GGUF files for local inference.
How does ornith-ai/Ornith-1.5-9B-GGUF handle long context?
ornith-ai/Ornith-1.5-9B-GGUF uses a hybrid attention design: only 8 of its 32 layers keep a full KV cache, and the rest use linear attention with a constant-size state. Long contexts cost far less memory than on a conventional model of this size.
How much VRAM do you need to run ornith-ai/Ornith-1.5-9B-GGUF?
You need about 7.96 GB of VRAM to run ornith-ai/Ornith-1.5-9B-GGUF entirely on the GPU using the Q6_K quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.
Can I run ornith-ai/Ornith-1.5-9B-GGUF on an 8 GB GPU?
Yes. With 8 GB of VRAM you can run ornith-ai/Ornith-1.5-9B-GGUF fully on the GPU using Q6_K (about 7.96 GB).
Can I run ornith-ai/Ornith-1.5-9B-GGUF on a 16 GB GPU?
Yes. With 16 GB of VRAM you can run ornith-ai/Ornith-1.5-9B-GGUF fully on the GPU using Q8_0 (about 10.04 GB).
Can I run ornith-ai/Ornith-1.5-9B-GGUF on a 24 GB GPU?
Yes. With 24 GB of VRAM you can run ornith-ai/Ornith-1.5-9B-GGUF fully on the GPU using BF16 (about 18.07 GB).
What context length does ornith-ai/Ornith-1.5-9B-GGUF support?
ornith-ai/Ornith-1.5-9B-GGUF supports a native context length of up to 262,144 tokens. A longer context grows the KV cache, so it increases the memory needed to run the model.
What is the best quantization for ornith-ai/Ornith-1.5-9B-GGUF?
For ornith-ai/Ornith-1.5-9B-GGUF, a strong default is Q4_K_M, which needs about 6.31 GB and keeps most of the quality while roughly halving the memory versus 8-bit. With VRAM to spare, Q5_K_M or Q6_K add a little more quality; if you are tight on memory, a smaller quantization still runs. Pick the highest quantization that fits your VRAM.