BA
Baidu
Multimodal model · China ·Jun 2025

ERNIE-4.5-VL-28B-A3B

Multimodal Open (Restricted)
28B
Parameters

About ERNIE-4.5-VL-28B-A3B

ERNIE-4.5-VL-28B-A3B is an AI model developed by Baidu, in the multimodal category, released in 2025, made available as an open-weights model with 28B params.

On this page you'll find ERNIE-4.5-VL-28B-A3B's full specifications. Review provider pricing and benchmark scores below, or compare ERNIE-4.5-VL-28B-A3B head-to-head with other multimodal models.

→ More models by Baidu → All multimodal models → AI benchmarks leaderboard

Links

Announcement ↗ Hugging Face ↗

Run ERNIE-4.5-VL-28B-A3B locally — what to buy

Share this build ↗

Sized to what ERNIE-4.5-VL-28B-A3B actually needs (~8k context at Q4), not to the biggest GPU: Good = cheapest that runs it, Better = best value, Best = most headroom. Speed is a hardware estimate; anything past ~120 tok/s is shown as instant.

Good
AMD Radeon RX 7900 XT
Consumer GPU · ~$899
~39.3 tok/s
Q4 · tight on context · up to 5k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →
Better
AMD Radeon RX 7900 XTX
Consumer GPU · ~$999
~47.1 tok/s
Q4 · fits full context · up to 8k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →
Best
NVIDIA GeForce RTX 5090
Consumer GPU · ~$1,999
~99.4 tok/s
Q4 · fits full context · up to 8k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →

Software support

·ExLlamaV2 ·KoboldCpp ·LM Studio ·MLX ·Ollama ·SGLang ·TensorRT-LLM ·llama.cpp ·vLLM ?KTransformers ?text-generation-webui

✓ measured · · compatible · — not supported. Informational only — speed is hardware-based.

Consumer GPU
12.6–99.4 tok/s
10 options →
Apple
15.7–41.8 tok/s
17 options →
Datacenter GPU
42.4–120+ tok/s
13 options →

Community

Would you run ERNIE-4.5-VL-28B-A3B again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

More models like ERNIE-4.5-VL-28B-A3B

AI
NeuroChess
· 72.3K
 BA
ERNIE 5.1
Baidu
 BA
ERNIE-5.0-Preview-1203
Baidu
 BA
ERNIE-5.0-Preview-1103
Baidu
 BA
ERNIE-5.0-Preview-1120
Baidu
 BA
ERNIE-4.5-VL-28B-A3B-Thinking
Baidu

Frequently asked questions

What is ERNIE-4.5-VL-28B-A3B? ▶

ERNIE-4.5-VL-28B-A3B is an AI model developed by Baidu, in the multimodal category, released in 2025. It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created ERNIE-4.5-VL-28B-A3B? ▶

ERNIE-4.5-VL-28B-A3B was developed by Baidu and released in 2025.

Is ERNIE-4.5-VL-28B-A3B open source or proprietary? ▶

ERNIE-4.5-VL-28B-A3B ships as an open-weights model: you can download and self-host the weights, though the licence may place some restrictions on use.

How much does ERNIE-4.5-VL-28B-A3B cost? ▶

Pricing for ERNIE-4.5-VL-28B-A3B depends on the provider. See the providers table on this page for the latest API rates.

How does ERNIE-4.5-VL-28B-A3B perform on benchmarks? ▶

Benchmark scores for ERNIE-4.5-VL-28B-A3B are listed on this page as they are published. You can compare it head-to-head with other models on the GenAIList compare tool.

ERNIE-4.5-VL-28B-A3B-Thinking : This LLM BEATS Gemini-2.5-Pro and GPT-5-High
ERNIE-4.5-VL-28B-A3B-Thinking | Advanced Visual AI Model | Baidu
How We Ran ERNIE-4.5-VL-28B-A3B-Thinking on Spheron AI and Solved Visual Reasoning Tasks
Baidu ERNIE-4.5-VL-28B-A3B: Vision Model for Special Tasks - Install and Test Locally
ERNIE-4.5-VL-28B-A3B-Thinking: A Breakthrough in Multimodal AI

Know a great video about ERNIE-4.5-VL-28B-A3B?