NameQuick 2.11.39 – Release Notes

Release date: 2026-05-26

New

Local model cleanup. You can now remove installed Ollama and MLX models from NameQuick, with a clear “Removing…” status while cleanup is in progress.
Smarter local model recommendations. NameQuick now groups local models by how well they fit everyday renaming, images, scanned PDFs, and advanced workflows, making it easier to choose the right Ollama, LM Studio, or MLX model.
Updated Gemini support. NameQuick now uses Google’s current Gemini Flash model by default and keeps Gemini-powered renaming aligned with the latest available API.

More reliable templates and custom prompts. Templates and custom prompts now handle dates, separators, suffixes, capitalization, and formatting instructions more consistently across cloud and local AI providers.
Better local AI model handling. Ollama, LM Studio, and MLX now detect model readiness, image support, thinking support, and request budgets more reliably before a rename starts.
Clearer local AI controls. Local AI settings now better explain which models are recommended, which models are basic, and when a model may not be the best fit for the selected workflow.
Safer model switching. NameQuick now handles local model changes more carefully, reducing surprise failures when switching between Ollama, LM Studio, or MLX models.
More helpful processing feedback. The Issues HUD and filename review details now give clearer guidance when a filename needs attention or a required detail is missing.
Improved subscription and activation handling. Trial, BYOK, and subscription states now update more reliably after activation, account changes, or launch.
More accurate Gemini cost estimates. Usage estimates now reflect the latest Gemini model pricing more closely.

Fixed deprecated Gemini preview model usage. Existing Gemini selections that point to older preview models now move to supported current models instead of failing when Google removes access.
Fixed custom prompts falsely reporting missing details. Required fields should no longer be marked missing just because the model used different capitalization, such as DATUM and Datum.
Fixed local AI filename failures with thinking-enabled models. Strict filename steps now avoid accepting thinking text as the final answer, reducing failures with Ollama, LM Studio, and MLX.
Fixed BYOK trial credit display. BYOK trial users should no longer see stale managed subscription balances while trial credits are still syncing.
Fixed unclear processing errors for custom prompts. NameQuick now does a better job explaining what to change instead of only showing a generic processing failure.