Fostera v0.8.1 — Multi-model synthesis

What's New

• Auto synthesis is now available on Core, bringing several AI perspectives together in one clear Soul response.
• Manual synthesis is now available on Premium, giving you control over each contributor and the final writer.
• Souls can keep working drafts for emails, texts, letters, notes, and direct messages, then continue revising them across turns.
• Synthesis now works with Soul abilities, including document and other creation workflows.

Improvements

• Synthesis details now show the plan, live progress, estimated and final Resonance, models used, and response details directly in the conversation.
• Auto synthesis chooses healthy, compatible providers while keeping its perspectives varied.
• Voice-transcription bars now make better use of the composer width on mobile and desktop.
• Model selection in Auto and manual modes is explained more clearly.
• Free users can preview synthesis; Auto is available on Core, and exact model-team selection is available on Premium.

Bug Fixes

• Synthesis now recovers more cleanly after temporary interruptions, with more reliable cancellation and retry behavior.
• Fixed older or canceled synthesis work occasionally replacing a newer revision.
• Resonance totals now accurately include the complete synthesis request.
• Soul proposal edits now return to chat more reliably after saving.