Reader - listen to any document: A full audiobook reader is now built in. Import a PDF, EPUB, TXT, or paste any text, and Reader speaks it sentence by sentence using Kokoro voices via Replicate. Tap any sentence to jump to it, pick your voice and speed, and your position is saved automatically. Accessible from the model picker alongside Chat and Voice. Requires your own Replicate API key (~$0.00088 per request). Spoken playback of any AI reply: Tap the play button next to any AI message to hear it read out loud. 30+ Kokoro voices, adjustable playback speed, full lock-screen and Control Centre controls, and an audio cache so the same message plays instantly the second time. Bring your own Replicate API key and generation costs go to your own Replicate credits (same provider & cost as Reader). Smaller install, on-demand local model: The app is now under 10 MB to install, down from over 700 MB. The on-device Llama 3.2 model has moved to an optional one-tap download from the model picker, so you only fetch it if you actually want offline AI. Existing local-model users: tap the Local LLM card to bring it back in a couple of minutes. Voice chat now powered by Google Gemini Live: We've moved voice chat from xAI to Google's Gemini Live for a faster, more natural realtime conversation. Many new voices to choose from - your Gemini API key now powers it - get yours free from Google AI Studio. Image input in voice chat: Share an image during a voice conversation - take a photo or screenshot and ask the assistant about it while talking. Powered by Gemini Live's multimodal input. Deeply customisable voice personality: The voice assistant's personality, accent, speech style, vocabulary, language preferences, and behaviour can all be shaped through the system prompt. Make it your own. Web search for OpenRouter chat: A Web Search toggle in the menu lets any OpenRouter model pull live information into its answers — current events, news, prices, weather, anything time-sensitive. This release also adds device-locale awareness, so users outside the US get region-relevant results (UK searches return UK sources, and so on). Charged to your OpenRouter credits at OpenRouter's standard rate. Cleaner model picker: Capability tags (Vision, Reasoning, parameter count, and more) now appear correctly on every OpenRouter model, including the "always latest" alias variants that previously showed up bare. Layout polish: Bigger thinking box for reasoning models. Wider message bubbles in landscape on iPhone and iPad. Plus the usual stability fixes throughout.

UI Bug fixes for 'Plus iPhones'

New in Version 2 Spoken playback of any AI reply. Tap play next to any message to hear it read aloud, 30+ voices, adjustable speed, lock-screen and Control Centre controls. Bring your own Replicate API key. 70× smaller install. The on-device Llama model is now a one-tap download from the model picker instead of being bundled with the app. Install drops from over 700 MB to under 10 MB. Existing local-model users: tap the Local LLM card to bring it back. Sharper landscape and rotation behaviour. Wider message bubbles in landscape on iPhone and iPad, larger thinking box for reasoning models.

This update is a major upgrade to the voice chat experience plus a powerful new web-search feature for chat. Voice chat now powered by Google Gemini Live We've moved voice chat from xAI to Google's Gemini Live for a faster, more natural realtime conversation. Many new voices to choose from, and your Gemini API key now powers it (get yours free from Google AI Studio). Show, don't tell — image input in voice chat You can now share an image during a voice conversation. Hold up a photo or screenshot and ask the assistant about it while talking. Powered by Gemini Live's multimodal input. Deeply customisable voice personality The voice assistant's personality, accent, speech style, vocabulary, language preferences, and behaviour can all be shaped through the system prompt. Set the kind of person it is, how it speaks, what units it uses, what topics it leans into — make it your own. This level of customisation wasn't possible with the previous voice provider. Web search for OpenRouter chat (BYOK) A new "Web Search" toggle in the menu lets any OpenRouter model pull live information into its answers — current events, news, prices, weather, anything time-sensitive. Toggle it on, ask, get an up-to-date reply. Charged to your own OpenRouter credits at OpenRouter's standard per-result rate. Plus the usual stability fixes and small improvements throughout.

• Conversation history now shows which AI model was used for each chat, with a one-tap option to delete all conversations at once. • Per-provider privacy notices for clearer data handling — including a dedicated on-device notice for the local model confirming nothing leaves your iPhone. • New Reasoning settings screen for compatible OpenRouter models — choose how much thinking the model does, from None all the way up to XHigh. • Cleaner, faster launch screen.

Online models now automatically understand your local date, time, and region Minor bug fixes

New: Thinking from reasoning models now displayed, see the models reasoning New: Long press to copy a message replaced with full share sheet

Bug fixes

Bug fixes

Screen now stays on for voice chats and for long AI text responses.

New: Real-Time Voice Chat Voice conversation powered by Grok - just like talking to a person 5 different voices: Ara, Eve, Rex, Leo, and Sal Unlimited storytelling - no 45-second cutoffs like other voice AI apps Full transcripts automatically saved after each voice session Dual system prompts for customized voice interactions xAI API Support Connect your xAI API key for Grok voice chat Get yours from x.ai Real-time voice conversation with sub-second response times Improvements: Voice conversations appear with microphone icon in chat history Text conversations appear with chat bubble icon Seamless switching between text and voice modes Enhanced iPad support with landscape orientation for voice chat