ERNIE Models by Baidu AI
About
ERNIE is a highly advanced family of large language models (LLMs) developed by Baidu, renowned for their proficiency in handling Chinese language tasks. The ERNIE family consists of several iterations, including ERNIE 1.0, ERNIE 2.0, ERNIE 3.0, ERNIE-Gram, and ERNIE-health, each offering diverse capabilities and model sizes. These models employ a transformer-based architecture and are pre-trained on vast amounts of data, encompassing textual corpora and knowledge graphs. This comprehensive pre-training equips ERNIE models with the ability to discern complex language patterns and structures, facilitating a broad spectrum of natural language processing applications, such as understanding and generating language. Notably, ERNIE 3.0 leverages knowledge graphs to enhance its comprehension and reasoning capabilities, while ERNIE Bot, powered by ERNIE 3.5, integrates plugins for real-time data retrieval and long text processing. These models have been embraced on multiple platforms, including Hugging Face's model hub, and have consistently delivered state-of-the-art performances on various benchmarks.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs reasoning, 64k context, and tool use.
Use when the workload needs reasoning, 32k context, and tool use.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| ERNIE X1.1 | Use when the workload needs reasoning, 64k context, and tool use. | 2025-09 | reasoning64k contexttool use | Current |
| ERNIE X1 Turbo | Use when the workload needs reasoning, 32k context, and tool use. | 2025-04 | reasoning32k contexttool use | Current |
| ERNIE 4.5 | Use when the workload needs 8k context. | 2025-03 | 8k context | Current |
| ERNIE X1 | Use when provider availability and model metadata match the workload. | 2025-03 | — | Current |
| ERNIE Speed Pro | Use when the workload needs 128k context. | 2025-01 | 128k context | Current |
| ERNIE Lite Pro | Use when the workload needs 128k context. | 2025-01 | 128k context | Current |
| ERNIE-4.0-8K | Use when the workload needs 8B parameters. | 2023-10 | 8B parameters | Current |
| ERNIE Bot | Use when provider availability and model metadata match the workload. | 2023-08 | — | Current |
Release Timeline
6 release groupsSpecifications(8 models)
| Model | Released | Context | Parameters | Vision | Multimodal | Reasoning | Fn Calling | Tool Use |
|---|---|---|---|---|---|---|---|---|
| ERNIE X1.1 | 2025-09 | 64k | — | Yes | Yes | Yes | Yes | Yes |
| ERNIE X1 Turbo | 2025-04 | 32k | — | Yes | Yes | Yes | Yes | Yes |
| ERNIE 4.5 | 2025-03 | 8k | — | No | No | No | No | No |
| ERNIE X1 | 2025-03 | — | — | No | No | No | No | No |
| ERNIE Speed Pro | 2025-01 | 128k | — | No | No | No | No | No |
| ERNIE Lite Pro | 2025-01 | 128k | — | No | No | No | No | No |
| ERNIE-4.0-8K | 2023-10 | — | 8B | No | No | No | No | No |
| ERNIE Bot | 2023-08 | — | — | No | No | No | No | No |
Available From(2 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| ERNIE Lite Pro | Baidu Qianfan | $0.03 | $0.059 | Serverless |
| ERNIE Speed Pro | Baidu Qianfan | $0.044 | $0.089 | Serverless |
| ERNIE X1 Turbo | Baidu Qianfan | $0.15 | $0.59 | Serverless |
| ERNIE X1.1 | Baidu Qianfan | $0.15 | $0.59 | Serverless |
| ERNIE X1 | Baidu Qianfan | $0.3 | $1.18 | Serverless |
| ERNIE 4.5 | Baidu Qianfan | $0.59 | $2.36 | Serverless |
| ERNIE 4.5 | Fireworks AI | $1.2 | $1.2 | Serverless |
Frequently Asked Questions
- What is ERNIE used for?
- ERNIE is used for reasoning, vision and multimodal work, and agent workflows and tool use. The family description and listed model capabilities point to those workloads as the best fit.
- How does ERNIE compare to ERNIE 4.5?
- ERNIE by Baidu AI is strongest where you need reasoning, while ERNIE 4.5 by Baidu AI is the closest related family to check for vision and multimodal work. ERNIE has 8 listed variants and reaches up to 128k context, while ERNIE 4.5 reaches up to 128k context, so compare the specs and pricing tables before choosing a production model.
- Which ERNIE model should I use?
- For the lowest listed input price, start with ERNIE Lite Pro through Baidu Qianfan at $0.03/1M input tokens. For the most capable/latest local choice, evaluate ERNIE X1.1 with 64k context and reasoning, tool use, function calling, and multimodal inputs.
