Compare

ERNIE 4.5 0.3B vs ERNIE 4.5 21B-A3B Thinking

Put two to four models side by side on the fields that decide a deployment: context, price, region, license, languages, hosting, and what is verified.

ERNIE 4.5 0.3BERNIE 4.5 21B-A3B Thinking
Provider BaiduBaidu
Parameters 0.4B21.8B
Context window Not verified: the provider has not documented this claim.Not verified: the provider has not documented this claim.
License apache-2.0apache-2.0
Modalities texttext
Release date 28 Jun 202508 Sept 2025
Languages Strong: en
Adequate: zh
Strong: en
Adequate: zh
Open weights yesyes
Tasks Low cost, Multilingual supportReasoning, Low cost

Where two models differ, this table shows the difference and stops. LLM EU publishes no winner badge without a documented score for that task, and a missing figure stays missing instead of becoming a zero.

When to use it

ERNIE 4.5 0.3B

Strengths

  • open weights are published on Hugging Face and can be downloaded without a vendor agreement
  • Apache-2.0 licence, so commercial self-hosting is permitted without a separate agreement
  • at 0.4B parameters it fits on a single accelerator
  • the model page lists en

Limits

  • catalogue only — LLM EU does not run this model, and nothing on this card implies an LLM EU endpoint
  • no evaluation has been run by LLM EU
  • the published configuration records no context window, so the card carries none
  • at 0.4B parameters, the weights fit on a single modest accelerator
  • a base checkpoint: it answers nothing useful without fine-tuning or a task head

Tasks

Low cost, Multilingual support

ERNIE 4.5 21B-A3B Thinking

Strengths

  • open weights are published on Hugging Face and can be downloaded without a vendor agreement
  • Apache-2.0 licence, so commercial self-hosting is permitted without a separate agreement
  • the model page lists en
  • instruction-tuned checkpoint with a chat template in the repository

Limits

  • catalogue only — LLM EU does not run this model, and nothing on this card implies an LLM EU endpoint
  • no evaluation has been run by LLM EU
  • the published configuration records no context window, so the card carries none
  • at 21.8B parameters, self-hosting needs a single accelerator with enough memory for the weights

Tasks

Reasoning, Low cost

Tasks: Low cost

Compare models

Compare models →