feat: add streaming display optimization option

[SameDesu123]

PR Checklist

• Required Checks
  • Have you added type definitions?
  • Have you tested your changes?
  • Have you checked that it won't break any existing features?
• If your PR uses models¹, check the following:
  • Have you checked if it works normally in all models?
  • Have you checked if it works normally in all web, local, and node-hosted versions? If it doesn't, have you blocked it in those versions?
• If your PR is highly AI generated², check the following:
  • Have you understood what the code does?
  • Have you cleaned up any unnecessary or redundant code?
  • Is it not a huge change?
    • This PR is a scoped, opt-in streaming display performance change.

Summary

Adds an experimental, opt-in Streaming Display Optimization setting for streamed assistant responses. The default off mode keeps the normal rendering path close to the existing behavior, while balanced and strong reduce UI churn during long streamed responses.

Changes

• Added a persisted streamingDisplayOptimizationMode setting with off, balanced, and strong options in Advanced Settings.
• Added labels and help text for all supported locales.
• Batched streaming display updates in balanced and strong modes with a timer plus requestAnimationFrame, while flushing pending output on completion or abort.
• Reduced active streaming message remount churn by keeping the row identity stable while performance mode is enabled.
• Computes streaming optimization eligibility in the parent chat list and passes plain props to the active streaming assistant message only.
• Added a session-limited parse cache in ChatBody for repeated rendering of long messages.

Additional Notes

• Validation: pnpm check reports 0 errors(in this changes).

Footnotes

1. Modifies the behavior of prompting, requesting, or handling responses from AI models. ↩

2. Over 80% of the code is AI generated. ↩

[kwaroran]

There seem to be too many effect hooks for a performance update.

Although we can expect better performance when this option is enabled when the update rate is high,

the PR adds a lot of hooks that run when the variable is updated even when the option is disabled, which can worsen performance when the option is disabled, or even if the update rate is not high enough with option enabled

[SameDesu123]

There seem to be too many effect hooks for a performance update.

Although we can expect better performance when this option is enabled when the update rate is high,

the PR adds a lot of hooks that run when the variable is updated even when the option is disabled, which can worsen performance when the option is disabled, or even if the update rate is not high enough with option enabled

Thanks, I think this concern is valid.

I’ll refactor this so the optimization state is computed once at the parent/current-chat level, and only the active streaming assistant message receives the optimized rendering props. When the mode is off, the normal rendering path should remain effectively unchanged.