v0.41.1

Full Changelog: v0.41.0...v0.41.1

v0.41.0

Full Changelog: v0.40.2...v0.41.0

v0.40.2

Full Changelog: v0.40.1...v0.40.2

v0.40.1

Full Changelog: v0.40.0...v0.40.1

v0.40.0

Full Changelog: v0.39.0...v0.40.0

v0.39.0

Full Changelog: v0.38.2...v0.39.0

v0.38.2

Full Changelog: v0.38.1...v0.38.2

v0.38.1

v0.38.1

Bug fixes

• REPL: Fixed invisible typed input — redrawFromCursor() now does a full line redraw via drawLine() instead of a partial redraw from line[e.pos:] which was empty when typing at end of line.

Enhancements

• WebUI: Theme toggle with localStorage persistence (🌙/☀️)
• WebUI: Model picker populated from GET /api/models with runtime model switching
• WebUI: Session rename via inline ✎ button → POST /api/sessions/:id
• WebUI: Thinking/reasoning content streamed live via IterationCallback
• WebUI: Copy buttons + collapse toggle on streamed assistant messages
• Server: GET /api/models returns available model profiles
• Server: POST /api/sessions/:id handles session rename
• API: Agent.SwitchModel() / Engine.SetModel() for runtime model hot-swap

Full Changelog: v0.38.0...v0.38.1

v0.38.0

Full Changelog: v0.37.22...v0.38.0

v0.37.22 — Default Model: deepseek-v4-flash

What's Changed

Default model switched: deepseek-chat → deepseek-v4-flash

• odek.go: defaultModel now deepseek-v4-flash (faster, cheaper, same 128K context)
• CLI: help text reflects new default
• Docs: API, README, CHEATSHEET examples all use v4-flash
• Fallback labels: all modelLabel fallbacks in main, repl, telegram updated

The v4-flash profile auto-configures:

• No extended thinking (faster/cheaper)
• 90s timeout
• 128K context window

Users can still use deepseek-v4-pro for deep reasoning (1M context, thinking enabled, 180s timeout).

Full Changelog: v0.37.21...v0.37.22

Full Changelog: v0.37.21...v0.37.22