Release date: 2026-04-28

New

• Added a more dependable local AI runtime package. The bundled local model setup now focuses on the text, image, and PDF workflows NameQuick uses, so local AI features can start with fewer optional components getting in the way.

Improvements

• Made local AI readiness easier to understand. NameQuick now separates setup, start, model, and runtime health states so the app can point you to the right next step.
• Made local model extraction more resilient. NameQuick can retry structured answers when a local model cuts off or repeats itself, while still rejecting malformed data.
• Improved update delivery for larger releases. App downloads and updates are now handled in a way that stays reliable as bundled local AI support grows.
• Preserved language choices more consistently. Managed AI requests now keep the selected output language through the full request.

Bug Fixes

• Fixed local AI startup blockers caused by optional media packages. NameQuick no longer lets unused audio, video, or speech components break document processing startup.
• Fixed unclear activation-limit errors. Account access screens now better explain when a license has reached its device limit and what to do next.
• Fixed duplicate processing queue entries. Files already queued or being added are now skipped more reliably, avoiding repeated work and confusing queue counts.
• Fixed unsafe structured responses from local models. Markdown-wrapped JSON, extra keys, unsupported envelopes, and malformed field shapes now fail clearly instead of becoming bad filenames.