KO

Kolibri 1

KolibriGGUF by Hob-forgeUpdated 2026-10-08
Download on Hugging Face
Parameters78B
Context256K
Layers50
Architecturekolibri1
Smallest quant28.6 GB
Quick answer

Kolibri 1 needs about 48.8 GB of memory at Q4_K_M (a 47.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 6 quants, from 28.6 GB (Q2_K) to 83.1 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 3.4 GB.

Kolibri 1 – 78 B Mixture‑of‑Experts reasoning model for German and English

Kolibri 1 is a 78 B parameter MoE model optimized for long‑context inference. It supports explicit reasoning and tool calling, with a native context length of 262 144 tokens. The model is released under Apache‑2.0 and requires a patched llama.cpp to run.

Good at

  • Long‑context reasoning
  • Tool‑aware generation
  • Efficient inference with MoE

Limits

  • Requires kolibri1 patch for llama.cpp
  • Large VRAM needed for full model
  • No vision support

Best for

  • Document summarization over long texts
  • Agentic workflows needing tool calls
  • Multi‑step reasoning tasks

Who it’s for. Developers building conversational assistants or internal knowledge systems that process German and English text. The model is suitable for users who can manage the memory requirements and have access to the patched llama.cpp runtime.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K28.6 GB 30.0 GB
Too big Get file →
Q3_K_M37.5 GB 38.9 GB
Too big Get file →
Q4_K_M47.5 GB 48.8 GB
Too big Get file →
Q5_K_M55.6 GB (2 parts) 56.9 GB
Too big Get file →
Q6_K64.2 GB (2 parts) 65.6 GB
Too big Get file →
Q8_083.1 GB (2 parts) 84.5 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (28.63 GB file, ~30 GB memory). For 12 GB VRAM, Q3_K_M (37.52 GB file, ~39 GB memory). For 16 GB VRAM, Q4_K_M (47.45 GB file, ~49 GB memory). Larger budgets can use Q5_K_M, Q6_K or Q8_0, noting the split shards for the latter three.

What GPU do you need for Kolibri 1?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) 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: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is Q2_K (28.6 GB file, about 30.0 GB in use).
  • 48 GB VRAM: the best quant that fits is Q3_K_M (37.5 GB file, about 38.9 GB in use).

Which Kolibri 1 quant should you choose?

Q4_K_M is the usual starting point for Kolibri 1: it balances file size and quality well. Here it is a 47.5 GB file.

The Q8_0 file is 83.1 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is Q2_K at 28.6 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturekolibri1
Parameters78B (approximate, mixture of experts)
Layers50
Trained context262,144 tokens
KV cache (fp16)100 KB per token

Source

Quantized byHob-forge
FamilyKolibri
Licenseapache-2.0
Downloads (HF)58,112
Last updated2026-10-08

Frequently asked questions about Kolibri 1

How much VRAM does Kolibri 1 need?

About 48.8 GB at Q4_K_M with an 8K context window. That is the 47.5 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 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 30.0 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Kolibri 1 GGUF?

It depends on the quant: Q2_K is 28.6 GB, Q3_K_M is 37.5 GB, Q4_K_M is 47.5 GB, Q5_K_M is 55.6 GB, Q6_K is 64.2 GB, Q8_0 is 83.1 GB.

What context length does Kolibri 1 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.4 GB at 32K.

Where can I download Kolibri 1 GGUF?

The files are on Hugging Face in the Hob-forge/Kolibri-1-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 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 on 8 GB VRAM?

Use the Q2_K quantization; it requires about 30 GB of VRAM, so 8 GB is insufficient.

Does Kolibri 1 support image inputs?

No, the model only processes text in German and English.

What context length can I use?

The native context is 262 144 tokens; longer contexts are possible but may increase latency.

Other GGUFs of Kolibri-1

Similar models

Browse all →

More from Hob-forge

All by Hob-forge →

More Kolibri models

All Kolibri →

Other large models

All large models →