v0.77.1

v0.77.1

• Updated OpenAI Chat Completions function calling to modern tools / tool_choice payloads.
• Kept legacy function_call response parsing as fallback compatibility.
• Added safer extraction helpers for assistant replies and function/tool calls.
• Disabled recursive tool-calling on second-pass model calls after external tool/API results are appended into chat context.
• Preserved no-temperature behavior for newer non-GPT-4 model families while keeping function/tool calling enabled.
• README changelog/version update.

v0.8.2 - Configurable OpenAI STT model for voice messages

v0.8.2 - Configurable OpenAI speech-to-text model support for Telegram voice messages

Changed

• Added configurable speech-to-text model selection through config.ini.
• Added support for gpt-4o-transcribe, gpt-4o-mini-transcribe, and whisper-1.
• Changed default voice transcription model from legacy whisper-1 to gpt-4o-transcribe.
• Kept backward compatibility with the existing EnableWhisper flag.
• Added OPENAI_STT_MODEL environment-variable fallback.
• Refactored voice message handling so the STT model comes from the main bot config object.
• Improved Telegram HTML safety by escaping transcribed text before formatting.
• Fixed duration checking so MaxDurationMinutes is treated as minutes while voice durations are handled as seconds.
• Improved voice transcription logging with Telegram user/chat/file/model metadata.
• Cleaned up voice handler registration in main.py.

Notes

Existing configs should continue working.

To select the STT model explicitly, add this to config.ini:

STTModel = gpt-4o-transcribe

Other useful values:

STTModel = gpt-4o-mini-transcribe
STTModel = whisper-1

v0.8.1

v0.8.1

• Added Perplexity API enabled/disabled status to the startup banner.
• Reads Perplexity status from [Perplexity] -> Enabled in config.ini.
• Kept startup/config handling simple and centralized in utils.py.

v0.8.0

v0.8.0

• Refactored Perplexity/Sonar helper module for the current Perplexity API workflow.
• Kept Perplexity on the OpenAI-compatible Chat Completions endpoint by default.
• Added configurable Perplexity endpoint support via [Perplexity] -> Endpoint.
• Kept sonar as the default Perplexity model.
• Added safer Perplexity config loading, retry handling, error handling, and response parsing.
• Preserved old Perplexity helper names/signatures for compatibility.
• Updated Perplexity-side OpenAI language detection helper for newer model payload rules.
• Removed stale hardcoded gpt-4o-mini fallback assumptions.
• Refactored DuckDuckGo search / agentic browsing helper for modern OpenAI tool-calling compatibility.
• Added modern tools / tool_choice / tool_calls support for the DuckDuckGo sub-agent.
• Kept backward-compatible legacy function_call parsing.
• Updated DuckDuckGo sub-agent payload rules for GPT-5.x-era/non-GPT-4 model families.
• Hardened DuckDuckGo fallback behavior when sub-agent calls, webpage fetching, or output parsing fail.
• Preserved non-agentic DuckDuckGo search behavior using lynx.
• Improved Telegram HTML safety and webpage dump handling.
• General cleanup of stale pre-5.x OpenAI model/default assumptions.