---

license: other

license_name: nvidia-open-model-agreement

license_link: https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-agreement/

base_model: Hcompany/Holotron4-30B-A3B

base_model_relation: quantized

pipeline_tag: image-text-to-text

library_name: gguf

tags:

- gguf

- llama.cpp

- nemotron

- nemotron-h

- mamba

- moe

- multimodal

- computer-use

- agent

---

Holotron4-30B-A3B GGUF

GGUF quantizations of Hcompany/Holotron4-30B-A3B, a 30B hybrid NemotronH MoE vision-language model (VLM) for Computer Use, tool-driven work, and agentic workflows.

For vision and audio input, use the included F16 multimodal projector (mmproj-Holotron4-30B-f16.gguf).

Upstream benchmarks

Holotron4 improves over its base model, Nemotron 3 Nano Omni, on GUI workflows and in environments with MCP tools, APIs, or code sandboxes. Gains are absolute percentage points.

| Benchmark | Interface | Nemotron 3 Nano Omni | Holotron4-30B-A3B | Gain |

| --- | --- | ---: | ---: | ---: |

| OSWorld | GUI | 21.0 | 76.3 | +55.3 |

| OSWorld 2.0 | GUI and code | 0.2 | 7.9 | +7.7 |

| AutomationBench | MCP | 19.4 | 35.6 | +16.2 |

| PinchBench | Terminal | 84.7 | 88.6 | +3.9 |

| ALE (Linux, code) | Terminal | 0.6 | 8.5 | +7.9 |

GGUF files

Due to the model's MoE intermediate dimensions (1856), standard K-quants (Q4_K, Q6_K) trigger internal fallbacks in llama.cpp to Q5/Q8 types. Therefore, clean 32-block quantizations (Q4_0 and Q8_0) are provided for optimal efficiency and accuracy.

| Quantization | File | Size | Notes |

| --- | --- | ---: | --- |

| Q4_0 | Holotron4-30B-Q4_0.gguf | 18.0 GB | Recommended balanced default (clean 4-bit) |

| Q8_0 | Holotron4-30B-Q8_0.gguf | 33.6 GB | Near-lossless high-precision reference |

| Multimodal projector | mmproj-Holotron4-30B-f16.gguf | 2.99 GB | Required for image and audio input |

Usage

Use a current llama.cpp build with the included chat template.

llama-cli \
  -m Holotron4-30B-Q4_0.gguf \
  -c 4096 -n 512 --temp 0.6 --top-p 0.95 \
  --jinja --chat-template-file chat_template.jinja \
  -p "<|im_start|>user\nWhat is computer use in AI?<|im_end|>\n<|im_start|>assistant\n"

For multimodal / vision and audio input:

llama-mtmd-cli \
  -m Holotron4-30B-Q4_0.gguf \
  -mm mmproj-Holotron4-30B-f16.gguf \
  --image screenshot.png \
  -p "Describe the interface in this screenshot."

Source

• Model: Hcompany/Holotron4-30B-A3B
• Chat template revision: 5218431
• License: NVIDIA Open Model Agreement
• Checksums: SHA256SUMS