Extensions/Plugins
text-generation-webui supports various plugins.
Enabling Plugins
In the Interface mode tab, you can enable or disable plugins. After switching, click Apply and restart the interface at the bottom to restart.
Built-in Plugins
Name | Effect | Explanation | Scene |
---|---|---|---|
api | API support | Creates an API with two endpoints, one for streaming at /api/v1/stream port 5005 and another for blocking at /api/v1/generate port 5000. | api |
google_translate | Translation of input and output | Translation | |
character_bias | Chat mode | In role-playing chat mode, adjust the character's state, such as the character's mood. | Role-playing |
gallery | Role-playing character gallery | Creates a gallery containing chat characters and their pictures. | Role-playing, NSFW |
silero_tts | Text-to-speech conversion | Uses Silero to convert the reply to speech. | TTS |
elevenlabs_tts | Text-to-speech conversion | Uses the ElevenLabs API to convert the reply to speech, requiring an ElevenLabs API key. | TTS |
whisper_stt | Voice input support | Uses whisper to support voice input and then converts the voice to text. | STT |
send_pictures | Image input support | Supports image input in chat mode using BLIP. | Image, multimodal |
multimodal | Text+Image input support | Supports text+image input in chat mode, requiring a model that supports multimodality. | Image, multimodal |
openai | openai API support | Supports openai-style APIs. | api |
superbooga | Support for input with very long context | Uses ChromaDB to create arbitrarily large fake context extensions, treating them as input text files, URLs, or pasted text. Based on superbig. | Multimodal, long context |
Other Plugins
If you need to install other third-party plugins in the community, download the plugin and copy it to the extensions directory under the text-generation-webui installation directory. Some plugins may also require environment configuration. Please refer to the corresponding plugin documentation for installation.
Name | Effect | Explanation | Scene |
---|---|---|---|
EdgeGPT | Uses Bing chat | Bing chat has internet access support | Bing, chat |
long_term_memory | Creates long-term memory for the robot | Stores the conversation record and vectorizes it so that the conversation robot can remember the previous conversation. | Memory |
complex_memory | Creates context memory for the robot | Provides additional background information for the robot to supplement the context of the conversation. | Context |
discord_bot | Sends responses to discord | Integrates with discord, allowing generated text to be sent to discord. | Discord |
oobabot | Implements chatbot on Discord | Integrates with Discord, allowing the chatbot to use text-generation-webui's capabilities for conversation. | Discord |
multi_translate | Enhances Google Translate functionality | Enhanced version of the google_translate extension, providing more translation options (more engines, saving options to file, instant on/off translation). | Translation |
api_advanced | Enhances api extension functionality | Provides interfaces similar to Kobold. Provides enhanced translation functionality, returning translated content in the response. | api |
barktts | Text-to-speech conversion | A simple implementation of Suno-AI's Bark text-to-speech, with implicit multilingual and simple sound support. | TTS |
webui_langchain_agent | Langchain integration, supports wikipedia search | Creates a Langchain agent that provides the ability to search wikipedia. | Langchain |