ERNIE 4.0 Reasoning: pricing, context window and benchmarks
ERNIE 4.0 Reasoning is a multimodal model from Baidu. It offers a 65,536-token context window, costs $12.20 per million input tokens and $55.24 per million output tokens, and averages 77.8 across the index's benchmark suite.
Pricing
ERNIE 4.0 Reasoning costs $12.20 per million input tokens and $55.24 per million output tokens. A blended 3:1 input-to-output mix works out to about $22.96 per million tokens. Cached input is listed at $0.99 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $12.20 | per 1M tokens |
| Output | $55.24 | per 1M tokens |
| Cached input | $0.99 | per 1M tokens |
| Blended 3:1 | $22.96 | estimate per 1M tokens |
| 1M in + 1M out | $67.44 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 81.6 | 131 | 78th | |
| GPQA Diamond | 75.0 | 145 | 76th | |
| SWE-bench Verified | 63.9 | 234 | 61th | |
| HumanEval | 82.3 | 179 | 70th | |
| MATH-500 | 79.1 | 231 | 62th | |
| AIME 2025 | 77.2 | 127 | 79th | |
| LiveCodeBench | 70.2 | 187 | 69th | |
| IFEval | 83.9 | 215 | 64th | |
| TruthfulQA | 81.2 | 148 | 75th | |
| HellaSwag | 86.7 | 187 | 69th | |
| MMMU | 75.1 | 161 | 73th | |
| Arena Elo | 1398.7 | 211 | 65th |
What ERNIE 4.0 Reasoning is good for
ERNIE 4.0 Reasoning is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $12.20 per million input tokens, it sits in the mid-range of the index, and the 65,536-token context window is enough for large files and multi-file diffs.
Strengths
- Accepts more than text, including image input where the provider supports it.
- Reasoning-oriented profile, useful for multi-step problems and structured analysis.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- Proprietary licence: you cannot download the weights or self-host the model.
- Premium input pricing at $12.20 per million tokens.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| ERNIE 4.0 Reasoning | Baidu | 65536 | $12.20 | 77.8 |
| MiniMax-Text Chat | MiniMax | 128000 | $4.26 | 77.8 |
| Sonar Max | Perplexity | 32768 | $11.60 | 77.8 |
| DeepSeek-Coder Instruct | DeepSeek | 32768 | $4.34 | 77.7 |
| GPT-5.1 Mini | OpenAI | 128000 | $12.19 | 77.7 |
| o4 Mini | OpenAI | 200000 | $11.92 | 77.7 |
FAQ
How much does ERNIE 4.0 Reasoning cost?
ERNIE 4.0 Reasoning is listed at $12.20 per million input tokens and $55.24 per million output tokens. Cached input is $0.99 per million tokens. A blended 3:1 input-to-output mix is about $22.96 per million tokens.
What is the context window of ERNIE 4.0 Reasoning?
ERNIE 4.0 Reasoning has a context window of 65,536 tokens and can emit up to 65,536 output tokens per request in this index.
Is ERNIE 4.0 Reasoning open weights?
No. ERNIE 4.0 Reasoning is a proprietary model with a proprietary licence, so you cannot download the weights.
How does ERNIE 4.0 Reasoning score on benchmarks?
ERNIE 4.0 Reasoning averages 77.8 across the scored benchmarks in this index, ranking #176 of 600 models. Its strongest relative result is Arena Elo at 1398.7.
What is ERNIE 4.0 Reasoning best for?
ERNIE 4.0 Reasoning is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $12.20 per million input tokens, it sits in the mid-range of the index, and the 65,536-token context window is enough for large files and multi-file diffs.