Skip to main content

​
What’s new

​
Platform

​
Audio support in AI integrations

AI integrations audio settings showing transcription and output options
AI integrations now supports audio capabilities with four new models:
  • gpt-4o-transcribe and gpt-4o-transcribe-mini for speech-to-text transcription
  • gpt-4o-audio and gpt-4o-audio-mini for audio output generation
Build applications that process or generate audio content—transcribe podcasts, create voice assistants, or add speech interfaces to your apps.

​
Web search in Agent automations

Agent delivering research summary via automation
Create automations that use web search without requiring API keys. Agent integrates LLM-native web search through AI integrations, enabling scheduled research, news summaries, or educational content delivery.

​
RulesSync support for replit.md

The open source project RulesSync now supports replit.md files. Keep your AI agent configuration files (CLAUDE.md, AGENT.md, and replit.md) synchronized across multiple projects.