AI Observatory / Model Radar Baidu / baidu/ernie-4.5-21b-a3b-thinking

Baidu: ERNIE 4.5 21B A3B Thinking

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.

131,072 Context
65,536 Max output
$0.07 / 1M Prompt price
2025-10-09 Created
01 / Snapshot

Pricing, context, modalities, and parameters.

Model Radar detail pages stay neutral and operator-readable: core metadata first, then workflow fit.

Provider Baidu Input modalities text
Output modalities text Prompt price $0.07 / 1M
Completion price $0.28 / 1M Request price N/A
Context length 131,072 Max completion tokens 65,536
Supported parameters frequency_penalty, include_reasoning, max_tokens, presence_penalty, reasoning, repetition_penalty, seed, stop, temperature, top_k, top_p
Best for baidu/ernie-4.5-21b-a3b-thinking

Baidu: ERNIE 4.5 21B A3B Thinking

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.

Deep analysis Coding workflows High-volume usage low-cost
03 / Colophon

Routes and exits.

Each model page stays simple: overview, compare, related models, then back to the public hub.