Compare · ModelsLive · 2 picked · head to head
ERNIE 4.5 21B A3B Thinking vs Gemini 2.5 Pro
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Gemini 2.5 Pro wins on 6/6 benchmarks
Gemini 2.5 Pro wins 6 of 6 shared benchmarks. Leads in math · knowledge · language.
Category leads
math·Gemini 2.5 Proknowledge·Gemini 2.5 Prolanguage·Gemini 2.5 Procoding·Gemini 2.5 Pro
Hype vs Reality
Attention vs performance
ERNIE 4.5 21B A3B Thinking
#64 by perf·no signal
Gemini 2.5 Pro
#116 by perf·no signal
Best value
ERNIE 4.5 21B A3B Thinking
36.1x better value than Gemini 2.5 Pro
ERNIE 4.5 21B A3B Thinking
325.7 pts/$
$0.18/M
Gemini 2.5 Pro
9.0 pts/$
$5.63/M
Vendor risk
Who is behind the model
Baidu
$45.0B·Tier 1
Google DeepMind
$4.20T·Tier 1
Head to head
6 benchmarks · 2 models
ERNIE 4.5 21B A3B ThinkingGemini 2.5 Pro
OpenCompass · AIME2025
Gemini 2.5 Pro leads by +12.5
ERNIE 4.5 21B A3B Thinking
76.2
Gemini 2.5 Pro
88.7
OpenCompass · GPQA-Diamond
Gemini 2.5 Pro leads by +24.5
ERNIE 4.5 21B A3B Thinking
60.2
Gemini 2.5 Pro
84.7
OpenCompass · HLE
Gemini 2.5 Pro leads by +14.6
ERNIE 4.5 21B A3B Thinking
6.5
Gemini 2.5 Pro
21.1
OpenCompass · IFEval
Gemini 2.5 Pro leads by +8.8
ERNIE 4.5 21B A3B Thinking
81.2
Gemini 2.5 Pro
90.0
OpenCompass · LiveCodeBenchV6
Gemini 2.5 Pro leads by +24.3
ERNIE 4.5 21B A3B Thinking
47.0
Gemini 2.5 Pro
71.3
OpenCompass · MMLU-Pro
Gemini 2.5 Pro leads by +15.0
ERNIE 4.5 21B A3B Thinking
70.8
Gemini 2.5 Pro
85.8
Full benchmark table
| Benchmark | ERNIE 4.5 21B A3B Thinking | Gemini 2.5 Pro |
|---|---|---|
OpenCompass · AIME2025 | 76.2 | 88.7 |
OpenCompass · GPQA-Diamond | 60.2 | 84.7 |
OpenCompass · HLE | 6.5 | 21.1 |
OpenCompass · IFEval | 81.2 | 90.0 |
OpenCompass · LiveCodeBenchV6 | 47.0 | 71.3 |
OpenCompass · MMLU-Pro | 70.8 | 85.8 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.07 | $0.28 | 131K tokens (~66 books) | $1.23 | |
| $1.25 | $10.00 | 1.0M tokens (~524 books) | $34.38 |