ERNIE-4.5-21B-A3B-Thinking is Baidu's upgraded lightweight MoE model, refined to boost reasoning depth and quality for top-tier performance in logical puzzles, math, science, coding, text generation, and expert-level academic benchmarks.
Model details →Baidu: ERNIE 4.5 21B A3B Thinking vs DeepSeek: DeepSeek V3 0324
Baidu: ERNIE 4.5 21B A3B Thinking and DeepSeek: DeepSeek V3 0324 are evenly matched on 6 axes — pick the one whose provider, latency, or licensing fits your stack.
DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team. It succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variet…
Model details →Side-by-side comparison
| Capability | Baidu: ERNIE 4.5 21B A3B Thinking | DeepSeek: DeepSeek V3 0324 | Winner |
|---|---|---|---|
| Context window Maximum number of input tokens the model can attend to in a single request. | 131K | 164K | 🏆 DeepSeek: DeepSeek V3 0324 |
| Input price (per 1M) Cost per million input tokens billed by the provider. | $0.07 | $0.20 | 🏆 Baidu: ERNIE 4.5 21B A3B Thinking |
| Output price (per 1M) Cost per million output tokens billed by the provider. | $0.28 | $0.77 | 🏆 Baidu: ERNIE 4.5 21B A3B Thinking |
| Tool / function calling First-class support for emitting structured tool calls. | — | Yes | 🏆 DeepSeek: DeepSeek V3 0324 |
| Vision input Accepts image inputs alongside text. | — | — | Tie |
| Reasoning mode Internal chain-of-thought / extended-thinking support. | Yes | Yes | Tie |
Frequently asked questions
Is Baidu: ERNIE 4.5 21B A3B Thinking better than DeepSeek: DeepSeek V3 0324?
Baidu: ERNIE 4.5 21B A3B Thinking and DeepSeek: DeepSeek V3 0324 are evenly matched on 6 axes — pick the one whose provider, latency, or licensing fits your stack.
What's the price difference between Baidu: ERNIE 4.5 21B A3B Thinking and DeepSeek: DeepSeek V3 0324?
Input: $0.07 vs $0.20 per 1M tokens. Output: $0.28 vs $0.77 per 1M tokens.
What context windows do Baidu: ERNIE 4.5 21B A3B Thinking and DeepSeek: DeepSeek V3 0324 support?
Baidu: ERNIE 4.5 21B A3B Thinking supports up to 131K tokens. DeepSeek: DeepSeek V3 0324 supports up to 164K tokens.
Do both Baidu: ERNIE 4.5 21B A3B Thinking and DeepSeek: DeepSeek V3 0324 support tool calling?
Baidu: ERNIE 4.5 21B A3B Thinking: not advertised. DeepSeek: DeepSeek V3 0324: yes.