Skip to content

v1.0.6 - Ultra-Fast Zero-Lag Streaming & Line Caching

Choose a tag to compare

View all tags
@PwnedBytes0x1 PwnedBytes0x1 released this 07 Sep 16:58
· 4 commits to main since this release
v1.0.6
e4454c9

Highlights in v1.0.6

  • Zero-Lag Streaming & Line Caching: Solved UI lagging behind network streaming during extended thinking and long output generation.
  • Delta Formatted Line Caching: Dynamic history line cache eliminates full-history ANSI re-parsing on each streaming token (from ~500-1800ms down to ~1.5ms per tick).
  • Windowed Live Thought Viewport: Renders active rolling tail during generation with line counter indicator, smoothly finalizing full unclipped thought block upon completion.
  • Debounced Refresh: Optimized streaming UI refresh interval to keep terminal lockstep with fast LLM network streams.
Assets 2
Loading