ERNIE 4.5 0.3B

ERNIE 4.5 0.3B needs an estimated 0.8 GB at Q4, 0.9 GB at Q8 or 1.3 GB at FP16. These figures are estimates based on the 361M parameter count, covering weights plus about 15% and 0.5 GB for a short context.

Yes. ERNIE 4.5 0.3B fits at FP16 on a 24 GB, 16 GB, 12 GB or 8 GB GPU, with an estimated 1.3 GB of VRAM.

Yes, at Q4 with an estimated 0.8 GB of VRAM. Assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits is 8 GB.

Yes. ERNIE 4.5 0.3B has open weights, so you can self-host it for full control of your data.

Run ERNIE 4.5 0.3B locally

No, ERNIE 4.5 0.3B only reads text. Its input modality is text, so images, audio, video and PDFs are not supported.

DiffuCoder 7B by Apple, Jamba Large by AI21 Labs and Gemma 3n E2B IT by Google DeepMind. They are matched on model kind and release date, with no comparable price.

DiffuCoder 7BERNIE 4.5 0.3B vs DiffuCoder 7BJamba LargeERNIE 4.5 0.3B vs Jamba LargeGemma 3n E2B ITERNIE 4.5 0.3B vs Gemma 3n E2B IT

Open weights. ERNIE 4.5 0.3B is released under the Apache 2.0 license.

Yes. ERNIE 4.5 0.3B is licensed under Apache 2.0, a permissive license that allows commercial use.

From the Hugging Face repository baidu/ERNIE-4.5-0.3B-PT.

Run ERNIE 4.5 0.3B locally

ERNIE 4.5 0.3B was made by Baidu and released on June 30, 2025.

More from Baidu

Yes. ERNIE 4.5 0.3B has open weights.