KO

Kolibri 1 APEX

KolibriGGUF by MyricUpdated 2026-10-09
Download on Hugging Face
Parameters78.1B
Context—
Layers—
Architectureimatrix
Smallest quant0.4 GB
Quick answer

Kolibri 1 APEX needs about 38.6 GB of memory at its main file (a 34.0 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 7 quants, from 0.4 GB to 45.0 GB (Q4_K).

Kolibri 1 APEX – 78.1B imatrix model for German/English text

Kolibri 1 APEX is a 78.1‑billion‑parameter imatrix model designed for German and English text. It supports long‑context reasoning and efficient inference. The model requires a forked llama.cpp to load.

Good at

  • Long‑context reasoning
  • Efficient inference
  • German/English text

Limits

  • Requires forked llama.cpp
  • Large memory for full context
  • No vision support

Best for

  • Document processing
  • Question answering over internal data
  • Agentic workflows with review

Who it’s for. Developers building conversational assistants in German and English that need long‑context and efficient inference. The model is intended for integration into systems where a human reviews the output before action.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.4 GB 1.0 GB
Fits Get file →
—24.4 GB 27.8 GB
Too big Get file →
—27.2 GB 30.9 GB
Too big Get file →
—34.0 GB 38.6 GB
Too big Get file →
IQ3_S35.0 GB 39.6 GB
Too big Get file →
—42.9 GB 48.5 GB
Too big Get file →
Q4_K45.0 GB 50.9 GB
Too big Get file →

For 8 k context, the 0.44 GB file uses about 1 GB of VRAM. If you need more context, the 34.95 GB IQ3_S quant uses about 39.6 GB, and the 44.98 GB Q4_K quant uses about 50.9 GB. Memory estimates are approximate.

What GPU do you need for Kolibri 1 APEX?

These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.0 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (27.2 GB file, about 30.9 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_S (35.0 GB file, about 39.6 GB in use).

Which Kolibri 1 APEX quant should you choose?

If memory is tight, the smallest file is the first one in the table at 0.4 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureimatrix
Parameters78.1B

Source

Quantized byMyric
FamilyKolibri
Licenseapache-2.0
Downloads (HF)2,846
Last updated2026-10-09

Frequently asked questions about Kolibri 1 APEX

How much VRAM does Kolibri 1 APEX need?

About 38.6 GB at its main file with an 8K context window. That is the 34.0 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Kolibri 1 APEX on an 8 GB GPU?

Yes, with the smaller files: the 0.4 GB file needs about 1.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Kolibri 1 APEX GGUF?

It depends on the quant: file is 0.4 GB, file is 24.4 GB, file is 27.2 GB, file is 34.0 GB, IQ3_S is 35.0 GB, file is 42.9 GB and 1 more.

Where can I download Kolibri 1 APEX GGUF?

The files are on Hugging Face in the Myric/Kolibri-1-APEX-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Kolibri 1 APEX use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Kolibri 1 APEX on 8 GB VRAM?

The 0.44 GB quant uses about 1 GB of VRAM at 8 k context, so it can run on 8 GB VRAM.

Does Kolibri 1 APEX support vision?

No, the model does not support vision.

Do I need a special build of llama.cpp?

Yes, you must use the brywil/llama.cpp kolibri1 branch to load the model.

Other GGUFs of Kolibri-1

Similar models

Browse all →

More from Myric

All by Myric →

Other large models

All large models →