ERNIE 5.1
Baidu · Undisclosed
Freshness through search grounding
ERNIE 5.1 ranks #1 in China and #4 globally on the LMSYS search board, with agent scores above DeepSeek-V4-Pro. Its multimodal edge is image-plus-live-information, not peak visual precision.
Last updated Oct 18, 2026Editorial scoring
Bottom line
Eight-dimension breakdown
editorial mapping · 0–100- ERNIE 5.188.1
Weights: perception 20 / documents 18 / charts 14 / video 14 / agents 14 / generation 6 / Chinese 6 / ecosystem 8. Definitions and sources on the methodology page.
Public benchmarks
as reported, not converted| Benchmark / metric | Value | Note |
|---|---|---|
| LMSYS 搜索榜 | #1 CN | #4 globally |
| Agent 能力 | > V4-Pro | Official claim |
Public scores come from different versions and prompts and are not interchangeable with our composite; they are one evidence input.
Scenario mapping
five stress tests, one scale| Scenario | Score | Observation |
|---|---|---|
| Chinese receipt OCR | 91 · Good on clean documents | |
| Chart data extraction | 88 · Stable on standard charts | |
| 30-min video QA | 83 · Limited long-video ability | |
| Mobile GUI control | 80 · Not a focus | |
| Chinese cultural context | 93 · Clear Chinese corpus advantage |
Strengths and limitations
Strengths
- Search grounding combines images with live information.
- Excellent Chinese formal writing and editing.
- Integrated with Baidu ecosystem (Drive, Wenku).
Limitations
- Video timing and GUI agents are weak.
- Loses detail in very long documents.
- Average on global sources and multilingual work.
Price and real cost
List price
¥8/M tokens
≈ ¥8 / 1M output tokens. For 1,000 images at 1,024px, ~1,100 image tokens each and 300 output tokens, the monthly cost is about ¥4.8 (assuming input at 25% of output price).
Use your volumeAbout ERNIE 5.1
Is ERNIE 5.1 open source?
Licence: Closed commercial. No public weights; access is via vendor API or an on-prem licence.
How much does ERNIE 5.1 cost?
≈ ¥8 / 1M output tokens. It ranks #11 with a composite of 88.1. Real bills depend on resolution, tiling, caching and volume discounts.
What is ERNIE 5.1 best for?
Policy and price image Q&A; Chinese content creation and editing; Search-grounded enterprise assistants。
Why does ERNIE 5.1 score 88.1?
The composite weights eight dimensions: perception 20%, documents 18%, charts 14%, video 14%, agents 14%, generation 6%, Chinese 6%, ecosystem 8%. Its strongest dimension is Chinese context & culture; its weakest is GUI agents & tool use.
Personal ranking · Not an official score
Every ranking, score and comparison on this page is an independent personal ranking, not an official score, and represents no vendor, institution or regulator. Scores are a subjective mapping of public evidence under our own weights — change the weights and the order changes. Disagree, or spotted a factual error? Contact the site owner: jizhi2024@foxmail.com — we check each report and log corrections in the changelog.