baidu models

4 baidu models from Baidu on Hugging Face, from 3.3B to 286.2B parameters, published by Baidu in BF16 and F8_E4M3, with contexts from 32K to 1M tokens. The smallest official model, Unlimited-OCR, needs about 7.5 GB of VRAM at its published precision; the cheapest live fit is RTX 3070 at $0.088/hr.

Pick a size

One row per official baidu size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.

ModelParametersNative VRAMFP8 VRAMINT4 VRAMLive GPU fit (native)Est. $/hrKV cache at 32K
Unlimited-OCR3.3B7.5 GB3.7 GB1.9 GBRTX 3070$0.088/hrnot published for this architecture
ERNIE-5.5-Flash-Base286.2B320 GB–160 GBAMD MI300X × 2$5.26/hr1.5 GB

VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.

Official models (4)

ModelParametersPublished asVRAM neededLive GPU fitEst. $/hr
Unlimited-OCR3.3BBF167.5 GBRTX 3070$0.088/hr
Unlimited-OCR-23.3BBF167.5 GBRTX 3070$0.088/hr
ERNIE-5.5-Flash-Base-BF16286.2BBF16640 GBRTX PRO 6000 × 7$9.63/hr
ERNIE-5.5-Flash-Base286.2BF8_E4M3320 GBAMD MI300X × 2$5.26/hr

VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.

More from Baidu

Start ComfyUI on a rented GPU.

© 2026 Aquanode. All rights reserved.

All trademarks, logos and brand names are the property of their respective owners.