How much RAM does Inkling need?

Inkling needs about 579 GB of memory at Q4_K_M quantization with 4K context. The download itself is 575 GB; the extra covers the context cache and runtime overhead.

Can I run Inkling on a laptop?

Not on typical consumer hardware. Inkling needs 579 GB, which is beyond a normal laptop or desktop.

Inkling: hardware requirements

Inkling needs about 579 GB of memory at Q4 quantization — a 575 GB download plus room for context and overhead. That is beyond consumer hardware; use it through an API instead.

Parameters

952B

Tier

Server only

Licence

See model card

Released

2026-07-16

Will it run on your machine?

8GB RAMNot enough memory 16GB RAMNot enough memory 24GB RAMNot enough memory 32GB RAMNot enough memory 64GB RAMNot enough memory 128GB RAMNot enough memory 8GB VRAMNot enough memory 12GB VRAMNot enough memory 16GB VRAMNot enough memory 24GB VRAMNot enough memory 32GB VRAMNot enough memory

Download sizes by quantization

Lower quantization means a smaller file and less memory, at some cost to quality. Q4_K_M is the usual starting point. What is this?

Quant	Download	RAM @ 4K	RAM @ 32K	Source
Q4_K_Mrecommended	575 GB	579 GB	587 GB	estimated
Q5_K_M	196 GB	200 GB	208 GB	measured
Q6_K	219 GB	223 GB	231 GB	measured
Q8_0	280 GB	285 GB	292 GB	measured
BF16	528 GB	532 GB	540 GB	measured

“Measured” sizes are read from published GGUF files (unsloth/Inkling-Small-GGUF). “Estimated” sizes are derived from the parameter count and are typically within a few percent. The RAM columns add the context cache and runtime overhead to the download size.

Model card on HuggingFace →API specs & pricing →Verified 2026-09-14Listed for 120 days after release