SophNet provides full-blood DeepSeek R1/V3, Qwen, QwQ models, and its own DeepSeek V3 Fast model, as well as text-to-speech, image OCR, document parsing, and other models.
Sophnet
SophNet provides full-blood DeepSeek R1/V3, Qwen, QwQ models, and its own DeepSeek V3 Fast model, as well as text-to-speech, speech-to-text, text embedding, image OCR, document parsing, and other models.
Configure
After installing the plugin, configure your Sophnet settings in the Model Provider section. This includes your API key (find it here) and Project ID. Save to use Sophnet.
Features
Large Language Models (LLM)
- Full access to DeepSeek R1/V3, Qwen, QwQ models
- Customizable parameters for each model
- Support for both chat and completion APIs
Speech to Text
- Convert audio files to text transcriptions
- Support for multiple audio formats: wav, mp3, m4a, flv, mp4, wma, 3gp, amr, aac, ogg-opus, flac
- File size limit: 1GB
- Audio duration: up to 5 hours
Text to Speech (TTS)
- Convert text to natural-sounding speech
- 14 different voice options in Chinese and English
- Adjustable parameters: volume, speech rate, pitch
- Streaming and non-streaming API support
- Multiple output audio formats
Text Embedding
- Generate vector embeddings for text
- Configurable dimensions: 1024/768/512/256/128/64
- Optimized for semantic search and text analysis
About Sophnet
Website Sophnet
Supported Models Models
Sophnet API Document API
Documentation
- See GUIDE.md for detailed usage, customization, and contribution instructions.
- Refer to code comments and docstrings for in-line documentation.
License
This project is licensed under the MIT License. See the LICENSE file for details.
Privacy
For information on data handling and privacy, please refer to PRIVACY.md.
Contributing
Contributions are welcome! Please read GUIDE.md for guidelines on how to get started.
Contact
For questions, issues, or feature requests, please open an issue on GitHub or contact the maintainers.