unsloth / image-text-to-text updated 1 month ago

Inkling-Small-GGUF

See Unsloth Dynamic 2.0 GGUFs for our quantization benchmarks. You can now run Inkling in Unsloth Studio with toggles for Thinking. Read our Inkling guide for analysis and instructions. See below for example of 1-bit UD-IQ1S GGUF running in Unsloth:

Params
Context
Downloads 30d
1.4 M
Likes
89
Commercial use: allowed apache-2.0 Not gated GGUF View on Hugging Face ↗

Download history

daily snapshots · 25 days
1.4 M855 K
Aug 27Sep 4Sep 12Sep 20

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

FileQuantSizeEst. VRAMVerdict on RTX 4090 · 24 GB
Inkling-Small-Q8_0-00005-of-00007.gguf Q8 48.6 GB 54.0 GB ❌ Won’t fit
Inkling-Small-UD-IQ2_XXS-00002-of-00003.gguf IQ2_XXS 49.4 GB 54.8 GB ❌ Won’t fit
Inkling-Small-UD-IQ3_XXS-00002-of-00003.gguf IQ3_XXS 49.4 GB 54.8 GB ❌ Won’t fit
Inkling-Small-UD-IQ2_M-00002-of-00003.gguf IQ2_M 49.5 GB 54.9 GB ❌ Won’t fit
Inkling-Small-UD-IQ1_M-00002-of-00003.gguf IQ1_M 49.7 GB 55.1 GB ❌ Won’t fit
Inkling-Small-UD-IQ1_S-00002-of-00003.gguf IQ1_S 49.7 GB 55.1 GB ❌ Won’t fit
Inkling-Small-UD-IQ3_S-00002-of-00004.gguf IQ3_S 49.7 GB 55.2 GB ❌ Won’t fit
Inkling-Small-BF16-00007-of-00011.gguf GGUF 50.0 GB 55.5 GB ❌ Won’t fit
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Licence
apache-2.0
First seen on the Hub
2026-07-30
Training datasets
undisclosed
Added to our catalog
2026-08-27
Compare with any image-text-to-text model