feat: add streaming LLM response support across all platforms

Cherry-picked from PR #828, resolved conflicts with main.
This commit is contained in:
teknium1
2026-03-11 08:56:37 -07:00
parent b2a4092783
commit 95d221c31c
6 changed files with 696 additions and 22 deletions

View File

@@ -219,6 +219,22 @@ compression:
# Options: "auto", "openrouter", "nous", "main"
# summary_provider: "auto"
# =============================================================================
# Streaming (live token-by-token response display)
# =============================================================================
# When enabled, LLM responses stream token-by-token instead of appearing
# all at once. Supported on Telegram, Discord, Slack (via message editing)
# and the API server (via SSE). Disabled by default.
#
# streaming:
# enabled: false # Master switch (default: off)
# # Per-platform overrides:
# # telegram: true
# # discord: true
# # api_server: true
# # edit_interval: 1.5 # Seconds between message edits (default: 1.5)
# # min_tokens: 20 # Tokens before first display (default: 20)
# =============================================================================
# Auxiliary Models (Advanced — Experimental)
# =============================================================================