XuanYuan Models by Du Xiaoman Data Intelligence
Last refreshed 2026-05-19. Next refresh: weekly.
Details
About
The XuanYuan family of large language models (LLMs), developed by Duxiaoman-DI, comprises several models of varying sizes and capabilities, all built upon the Llama architecture and trained on extensive Chinese and English datasets. These models are specifically designed for financial applications, exhibiting superior performance in tasks like financial analysis, question answering, and text generation 347. A notable feature is their ability to manage long context lengths, with some models handling up to 16k tokens 5. The family includes both base models and fine-tuned chat models, with quantized versions available to enhance efficiency 268. These models can be accessed on platforms like Hugging Face, among others.
Decision facts
- Best fit
- coding
- Capability starting point
- XuanYuan 176B
- Lowest tracked input
- Not tracked
- Closest related family
- XuanYuan 2
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| XuanYuan 176B | Use when the workload needs 176B parameters. | 2024-02 | 176B parameters | Current |
| XuanYuan 70B | Use when the workload needs 70B parameters. | 2024-02 | 70B parameters | Current |
| XuanYuan 13B | Use when the workload needs 13B parameters. | 2024-02 | 13B parameters | Current |
| XuanYuan 6B | Use when the workload needs 6B parameters. | 2024-02 | 6B parameters | Current |
Release Timeline
1 release groupSpecifications(4 models)
| Model | Released | Parameters |
|---|---|---|
| XuanYuan 176B | 2024-02 | 176B |
| XuanYuan 70B | 2024-02 | 70B |
| XuanYuan 13B | 2024-02 | 13B |
| XuanYuan 6B | 2024-02 | 6B |
