A plain-language record of what's changed in IWV Chat AI with each release — what we fixed, what we improved, and what's new. The same information is available inside the app under Help → Release Notes & Errata.
Added: The Help menu now has a Report AI Content entry. You could already report any AI response from the flag on the response itself; this new entry lets you report AI content that is no longer on screen — for example from a conversation you have since deleted. As always, you choose a category, see the exact report text before anything happens, and nothing is sent unless you send it from your own email app.
Fixed: On Android, pressing the Back button while the Report AI Content window was open could close the wrong thing. Back now closes the report window itself.
Changed: When you finish dictating, the microphone window’s main button now sends your words straight to the assistant. It reads “Stop & Send”. If you would rather review or edit first, turn off “Send straight to the assistant” in Settings → Voice and the button goes back to only filling in the message box. Either way your words are put in the message box first, so nothing is lost if the assistant is busy. Your words are never sent on their own: if dictation ends because you went quiet for a minute, or because the microphone stopped working, the text is placed in the message box and left there for you.
Fixed: After using and leaving hands-free Conversation Mode, the microphone button could disappear, leaving no way to dictate for the rest of the session. Anything you typed yourself is never touched.
Changed: The list of downloadable speaking voices on desktop is now six English voices whose licenses could be fully verified (Kristin, LJSpeech, Bryce, John, Norman, and Cori). Voices you have already downloaded keep working and stay selectable; they just cannot be downloaded again from within the app. The built-in default voice is unchanged.
Improved: Internal release-quality hardening across these versions — build-time integrity checks that now stop a build immediately on any failure, and verification that the voice list matches the license-verified set exactly. No other changes to how the app looks or behaves.
Added: When a reply is going to take a while, the app now says so: a long message to a slow model shows an up-front time estimate, and any request that runs past 20 seconds without a response shows how long it has been working. Cancel remains available throughout.
Fixed: Long conversations no longer get refused with “Message too large” when they would actually have fit — the size check over-estimated by up to 80%; the AI engine itself now has the final say. If a reply really is cut short by a full memory, the app says so clearly and suggests starting a new chat.
Fixed: Cancelling voice input while your speech was still being transcribed now actually stops the transcription. Rapidly switching AI models could leave the app confused about which engine was running; switches are now tracked correctly. Very occasionally the AI would deliver an entire answer formatted as a block of code — such answers now display, and read aloud, as normal text.
Changed: On desktop, new installations now start with the Fast AI model — the same default as phones and tablets. It is a quarter of the size of the previous desktop default and can also answer questions about images. If you already picked a model, your choice is kept.
Improved: AI model and speech-recognition downloads are now fetched from fixed, verified versions of their sources, so a file changing upstream can never alter what you receive. All confirmation dialogs now use the app’s own dialogs — on some Linux desktops the old system dialogs could open hidden behind the window. On Linux, closing the app from the system (shutdown or log-out) now saves the conversation in progress before exiting.
Versions 1.20.7 through 1.27.2 were internal build revisions and were not released. Their user-relevant changes are collected here.
Added: Your conversations and logs now live in a folder you can find: Documents / IWV_Chat-AI. Existing conversations move there automatically, and you can put the folder anywhere you like from Settings. AI models stay where they were, so moving the folder does not move gigabytes around.
Added: The Fast AI model has been replaced with a newer one that is also able to see images, so both bundled models can now answer questions about a photo.
Added: Tapping the microphone now opens a dictation window with your words appearing live, a level meter so you can see you are being heard, and a Stop button. You can also choose which speech-recognition model to use, and the app will warn you before a session if the one you selected is slower than your speech.
Added: A context meter appears above the message box once a conversation starts filling the AI’s memory, at 75% and again at 90%, so you can start a new chat before older messages begin dropping out. Attached images now count towards that total. Assistant replies also gained a Read aloud button, and models that show their reasoning display it in a collapsed section you can expand.
Fixed: Opening the camera or microphone on desktop could fail while the AI engine was loaded, because both wanted the graphics card at once — the engine now steps aside and restarts afterwards. On Android, the pause between spoken sentences now actually works. Text inside code blocks is no longer invisible in the Light theme.
Improved: Sending is more reliable (failed sends retry, the app waits for the engine to finish loading, and a reply stuck repeating itself is cut off). The engine no longer reloads after a pause in conversation. Downloaded AI models are verified before use, with a Verify Installed Models option for existing downloads. Bug reports no longer include your username or any email address. Text you paste is cleaned of hidden characters before it reaches the AI.
Improved: Hands-free Conversation Mode now lets you speak for up to 30 seconds per turn, up from 12, and live transcription stays equally responsive no matter how long you speak. On desktop, your words now appear on screen as you speak them (a running estimate that may correct itself), the “Listening chunk period” setting now controls how quickly they appear, and the app waits until the microphone is actually recording before showing “Listening” so your first words are no longer missed. It is also better at telling a short pause mid-sentence from the end of your turn, especially with noise-reducing microphones.
Fixed: Choosing a second image while a previous one was still being prepared could keep the wrong image. Asking about an image without the required AI model now opens the AI Models screen so you can watch the download progress. Settings, Help, and the setup wizard now scroll properly on short windows.
Improved: AI responses are cleaned of hidden characters that could be copied along with the text without being visible on screen.
Improved: Release builds now use code shrinking and optimization, reducing the app’s download and install size (1.19.8). The status window shown while the AI is working now matches your selected theme instead of always appearing as a dark card (1.19.2).
Added: The Help button opens a Help window with the same buttons and layout as Settings on every platform (1.19.1). The Android back button closes an open menu or dialog instead of leaving the app, the app only exits when nothing is open, and on desktop the Esc key closes the topmost dialog (1.19.0).
Versions 1.18.x and 1.19.3–1.19.7 were internal build revisions and were never released.
Fixed: In Conversational Mode, if you press Done Speaking (or your pause is detected) and no words could be understood, the app now says so out loud instead of silently listening again — previously a microphone problem could look like the app simply ignoring you. Quiet waiting periods are unaffected.
Fixed: On Android, the status display now updates to “Working…” immediately when your speech is being processed, instead of appearing frozen on “Listening…” until the answer was ready.
Added: Conversational Mode on Android now detects when you stop speaking, just like on desktop — your words are processed after your natural pause instead of waiting for the 12-second listening window or the Done Speaking button (both still work as backups). Detection adapts to your device’s microphone and background noise, and calibrates itself to your voice over the first few turns.
Improved: Voice transcription on Android now runs only on the phone’s performance cores instead of all cores. On devices that mix fast and efficiency cores, this makes speech-to-text noticeably faster — the same optimization the AI engine already used.
Fixed: On Android, the AI model could be loaded into memory twice when two parts of the app warmed the engine at nearly the same moment (for example, app startup and entering Conversational Mode). The duplicate copy could not be freed, roughly doubling memory use — on lower-memory devices this could make the microphone go silent during voice conversations or cause the system to close the app. Model loading is now strictly one-at-a-time, and a second request simply reuses the load already in progress.
Fixed: The app icon could show black corners around the circular logo instead of transparency — on Windows (a malformed icon file that some renderers mis-parsed) and on Android (launcher icons exported with opaque black corners). All icon assets regenerated with verified transparency at every size.
Improved: Reporting an AI response now always includes the correct originating question and notes any image that was part of that exchange — including follow-up answers about an image attached earlier in the conversation (marked as such). Only attachment names and types are included, never the files themselves.
Improved: Very large images (30 MB and up) are now processed with a memory-safe method adopted from NotesXML: the image is decoded and scaled directly, without building huge intermediate copies — large panoramas and scans attach faster and more reliably.
Fixed: Images added by drag-and-drop or paste were being scaled to a smaller size than the vision model expects; they now use the correct per-model resolution (matching images added via the file picker).
Fixed: Attaching a large image could fail silently — nothing appeared and no reason was given. Failures now post a notice directly in the chat, and the image size limit has been raised from 10 MB to 100 MB (images are scaled to the AI model’s input size regardless, so ordinary large photos, panoramas, and scans now simply work).
Fixed: Spoken replies could clip the first word or two when speech started after a quiet period; a brief silent lead-in now wakes the audio output first, independent of your pause setting.
Improved: Much shorter gaps between spoken sentences: the voice engine now stays loaded for the whole session instead of restarting for every sentence, and the between-sentence pause setting now controls exactly that pause and nothing else.
Fixed: The AI engine was being launched with a far larger context window than intended, allocating a very large memory cache that slowed response generation — most visibly on integrated-GPU systems (roughly 4x slower than the hardware can deliver). The intended context size is now enforced.
Added: The AI engine now runs on Intel Arc/Iris integrated GPUs on Linux (previously CPU-only on those systems), with roughly 3x faster responses on tested hardware. The underlying inference engine was updated accordingly.
Fixed: Images sent to the AI were being downscaled below the vision model’s native input size, reducing OCR and image-understanding quality. Vision inputs now use the correct per-model resolution.
Fixed: On systems where the AI was set to CPU-only, background GPU activity could still occur and slow the machine; CPU-only now fully disables the GPU path.
Fixed: Live voice-activity sampling could read stale audio on systems where the recorder writes in large bursts, disabling automatic end-of-speech detection. Recording is now written at fine granularity and only fresh audio is analyzed.
Changed: Voice Settings now apply immediately as you change them. The Save button has been removed — previously it sat below the fold and changes were silently discarded if it was missed.
Added: The in-app Release Notes & Errata page.
Added: Modulation-based voice detection for microphones whose processing (automatic gain control / noise suppression) flattens loudness so that level-based detection cannot work. The right detector is selected automatically per machine.
Fixed: Voice detection could lock into a “speech” state under steady background noise (for example a cooling fan), preventing automatic end-of-speech; the adaptive noise floor was redesigned (1.7.0).
Fixed: The inactivity timeout could end Conversational Mode while an utterance was still being processed, discarding it (1.7.1), and could end an active conversation after 2.5 minutes on systems without working end-of-speech detection (1.7.3).
Added: Per-machine speech-level auto-calibration for voice detection (1.7.2).
Fixed: Silence or background noise could occasionally be transcribed and sent to the AI as a phantom turn (1.6.0); an over-aggressive background-audio guard could exit the mode and discard a valid spoken question (1.6.2); idle sessions previously never timed out (1.6.2).
Added: Conversational Mode safeguards: entry requires a deliberate mouse/touch press; opening the camera or typing exits the mode; continuous background audio (TV, music) no longer keeps it running indefinitely (1.6.1). The AI engine is warmed when entering the mode, making the first spoken answer much faster (1.6.0).
Fixed: Text-to-speech could speak formatting symbols (“asterisk asterisk”) and wasted synthesis on punctuation-only fragments (1.5.1).
Added: Persistent speech-recognition engine: the transcription model loads once per session instead of once per utterance, removing a fixed delay from every voice turn (1.5.0).
Fixed: End-of-speech detection previously depended on transcription timing, so the “pause after you stop speaking” setting did not measure your actual pause, and long utterances could be cut off mid-speech.
Added: Real-time voice-activity detection: the pause setting now does exactly what it says.
Added: Voice model selection with a downloadable catalog of additional voices, multi-speaker voice support, and a Test button (1.3.0–1.3.1).
Fixed: The spoken “Processing” cue delayed the AI request instead of playing alongside it (1.3.5); light-theme title bar contrast (1.3.4); settings styling issues (1.3.2–1.3.3).
Added: Conversational Mode tuning controls: adjustable pause after you stop speaking, adjustable pause between spoken sentences, adjustable live-transcription chunk period.
Added: Android 15 compatibility (16 KB memory-page devices) and updated target SDK (1.1.0).
Changed: Application identifier finalized for store distribution (1.1.1); launcher icon corrections (1.1.2); AI model catalog update (1.1.4); desktop GPU backend housekeeping (1.1.5).
This page currently covers releases 1.1.0 through 1.17.1; notes for the most recent releases are being finalized. Have a question about a specific fix? Contact support.