v0.3.2

What's Changed

The token limit for Claude's dialogue history has been removed.

Note: when the total number of tokens in the dialogue history exceeds 9216, Claude's responses may become irrelevant to the previous context.

Changelog: v0.3.1...v0.3.2

v0.3.1

• Improved bot response when mentioned in groups along with other bots.

Changelog: v0.3.0...v0.3.1

v0.3.0

Changelog

• Discarded invitation mode.
• Maintain connection with both Claude and Bard, allowing users to continue previous conversations after switching mode back.
• Users can leave the API field for Claude blank, in which case the bot defaults to Bard mode, and vice versa.

v0.2.4

Changelog: v0.2.3...v0.2.4

v0.2.3

What's Changed

• Asynchronously request bard's response to avoid blocking
• Set default streaming cutoff to 50

v0.2.2

What's Changed

• Enable concurrent_updates for multiple users
• Add error_handler for better logging output

v0.2.1

Features

• Support of official Claude API and reverse engineered Bard API (command: /mode to switch between them)
• Claude's streaming output (command: /cutoff to adjust the frequency of streaming, defaults to 100)
• Modify Claude's model (defaults to v1.3) and temperature (defaults to 1) (commands: /model and /temp)
• Show reference links and Google search keywords from Bard's answers (button: 🔍 Google it)
• Switch between different drafts provided by Bard's answers (button: 📝 View other drafts)
• Support of partial Markdown
• Private chat, group chat, invitation mode, independent chat session