Downloads · 30 days
119
4% of all-time downloads
batiai/Nemotron-3-Nano-Omni-30B-Text-Only-GGUF
Nemotron-3-Nano-Omni-30B-Text-Only-GGUF is a text generation model from batiai. Use it when you need the model to write or continue text. It is set up for llama.cpp. The card lists the license as other.
<p align="center" <a href="https://flow.bati.ai"<img src="https://img.shields.io/badge/BatiFlow-macOS%20AI%20Automation-blue?style=for-the-badge&logo=apple" alt="BatiFlow"</a <a href="https://ollama.com/batiai/nemotro…
Downloads · 30 days
119
4% of all-time downloads
All-time downloads
2.7K
Public
Repo size
124 GB
Likes
0
Public
Click a slice to open those files.
.gguf61.9 GB · 100%
From the Hugging Face model README
Text-only extraction from NVIDIA's Nemotron 3 Nano Omni Reasoning multimodal model. The text backbone (NemotronH MoE 30B-A3B) extracted, BF16 → IQ3/IQ4/Q5 quantized. Built and verified by BatiAI for BatiFlow.
The original Nemotron 3 Nano Omni is a multimodal reasoning model with:
This GGUF release contains only the text backbone because llama.cpp does not yet support the Omni multimodal architecture. The vision and audio encoders are stripped; the text backbone (which is a standard NemotronHForCausalLM) is converted as-is.
Get the multimodal version when available: We are watching for upstream llama.cpp support and will publish the full Omni (vision+audio+text) GGUF as soon as it lands.
ollama pull batiai/nemotron3-nano-omni-text:iq4
| Quant | Size | VRAM target | Recommended For |
|---|---|---|---|
| IQ3_XXS | 17GB | ~20GB | 24–32GB Mac |
| IQ4_XS | 17GB | ~20GB | 32GB Mac (recommended) |
| Q5_K_M | 25GB | ~28GB | 36GB+ Mac (highest quality) |
| Your Mac RAM | IQ3_XXS (17GB) | IQ4_XS (17GB) | Q5_K_M (25GB) |
|---|---|---|---|
| 16GB | ⚠️ Heavy swap | ⚠️ Heavy swap | ❌ |
| 24GB | ✅ | ✅ | ❌ |
| 32GB | ✅ Fast | ✅ Recommended | ⚠️ Tight |
| 36GB+ | ✅ | ✅ | ✅ Best quality |
Both share the same NemotronH MoE 30B-A3B backbone, but the Omni variant is reasoning-tuned:
| Nemotron 3 Nano | Nemotron 3 Nano Omni (text) | |
|---|---|---|
| Tuning | General agentic | Reasoning-focused |
| Step-by-step output | Standard | Stronger |
| Tool calling | ✅ | ✅ |
| Multimodal | — | ⚠️ Stripped (vision/audio not in this GGUF) |
If you want vision/audio, watch this repo — multimodal release pending llama.cpp support.
language_model.* tensors and re-saves with the standard NemotronHForCausalLM configgeneral.author=BatiAI)language_model.* tensors → standard NemotronHForCausalLM)imatrix --chunks 200 calibratedBatiFlow — free, on-device AI automation for Mac. 5MB app, 100% local, unlimited. 57+ built-in tools for calendar, notes, reminders, files, email, browser, messaging.
Quantized from nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16. License: NVIDIA Open Model License.