Skip to main content
PromptQuorumBuilt for humans. Structured for AI.
Home/Power Local LLM
Mobile & Edge LLMs

LLM Studio Go Review: Local AI Chat, Voice and Music on iPhone and iPad

·7 min readBy Hans Kuepper · Founder of PromptQuorum · Discovery engine for open-weight & open-source AI

LLM Studio Go (version 1.2.1, released on 22 October 2026) is a free iPhone and iPad app that, per its App Store listing, runs chat, image, voice, music and speech models on the device; it shows ads, and its App Store privacy labels list tracking identifiers. It is a different product from LM Studio. Every statement here comes from the listing and the developer's policy; PromptQuorum has not tested the app.

Key Takeaways

  • Maintenance: the listing appeared on 15 October 2026 and four versions (1.0.8, 1.0.9, 1.2 and 1.2.1) were released between 16 and 22 October 2026, so the app is active but one week old; no stale-project warning applies.
  • Publisher: a single named developer, Yevhenii Svistunov; no company is named.
  • Track record: the App Store showed no ratings yet at the time of checking.
  • Basis: the App Store listing in nine storefronts, the version history, the privacy policy and the support page, read on 24 October 2026, with no hands-on testing.

📍 In One Sentence

LLM Studio Go runs language, image and audio models on the iPhone or iPad itself, free of charge and paid for by ads.

💬 In Plain Terms

You download a model once, then chat, talk, generate images or make music without an account; the models are separate downloads, mostly from Hugging Face.

📌Note: PromptQuorum has not installed or tested LLM Studio Go. The feature, speed and privacy statements below are the developer's own.

Read in:
🇺🇸en🇩🇪de🇫🇷fr🇯🇵ja🇨🇳zh🇪🇸es🇧🇷pt🇸🇦ar🇰🇷ko

LLM Studio Go is a free, ad-supported iPhone and iPad app from the developer Yevhenii Svistunov that downloads open models in GGUF and MLX format and runs chat, image, voice, music and speech generation on the device. It is not LM Studio, the desktop app from lmstudio.ai. This review reflects the App Store listing in nine storefronts, the version history and the developer's privacy policy and support page, checked on 24 October 2026; PromptQuorum has not tested the app.

LLM Studio Go Review: Local AI Chat, Voice and Music on iPhone and iPad

What Is LLM Studio Go?

LLM Studio Go is an iPhone and iPad app that downloads open models in GGUF and MLX format and runs them on the device for chat, image generation, voice conversation, music and text-to-speech. The App Store listing names Yevhenii Svistunov as the publisher; the developer site it links, vans.wtf, is a personal portfolio headed "Eugene Svistunov".

It is not LM Studio. LM Studio is a separate desktop application (review) whose iPhone app is called Locally, according to its changelog; the listing and the developer's pages for LLM Studio Go state no connection to it. The developer's site also shows an Android app named LLM Studio (Google Play, package com.vans.llmstudio, the same identifier as this app's iOS bundle ID); that is a separate listing and is not covered here.

Get It

Install it from the App Store on an iPhone or iPad with iOS or iPadOS 18 or later; the AI models are a separate download inside the app.

Platform
Where to get it
iPhone / iPadApp Store
Websitevans.wtf
Privacy policyPrivacy policy (iPhone and iPad)
SupportSupport page
Source codeNo repository found

Version reviewed: 1.2.1, released on 22 October 2026 per the App Store version history, which is the latest upstream update at the time of checking, so no stale-project warning applies. No dedicated product website was found: vans.wtf is the developer's portfolio, which showed only the Android app when checked, and it hosts the iOS privacy policy and the support page.

Pricing and Ads

The app is free to download and every local AI feature works without paying; ads can appear in the model catalog, the model picker and the audio-creation screen, and a one-time purchase called Remove Ads Forever hides them.

Storefront
Remove Ads Forever
United States$9.99
United Kingdom£9.99
Germany, Spain, France9,99 €
Japan¥1,500
BrazilR$ 59,90
South Korea₩17,000
Saudi ArabiaSAR 39.99

Prices as shown on each App Store storefront on 24 October 2026, not converted. The purchase is not a subscription and can be restored with the same Apple Account, per the policy. The app was not found in the China storefront.

Features Confirmed by the Sources

Every item below comes from the App Store description, the release notes or the privacy policy; none has been tested.

  • Chat and files. Chat with supported language and vision models and attach PDF, Word, Excel, PowerPoint, text files or images to a conversation.
  • Model library. Browse GGUF and MLX models, import your own files and manage downloads; the policy names Hugging Face as the download source, and the app can measure how fast a model runs on your device.
  • Images. Generate images with compatible diffusion models; the notes for version 1.2.1 say devices with older graphics hardware now switch to the processor automatically for more stable results.
  • Audio and voice, new in 1.2. Speak to a chat model and hear the answer with live captions, with recognition, reply and speech all on the device; create music or sound effects of up to 2 minutes, songs with vocals from your own or a model-written text, and text-to-speech in more than 30 languages, with WAV export. Songs need an iPhone or iPad with at least 12 GB of RAM and are made while the app is open.
  • Optional web search and local API. DuckDuckGo search and link reading are off by default; an OpenAI-compatible API for devices on a trusted local network is optional and also off by default.

Privacy, Ads and Tracking

The privacy policy (effective 18 October 2026, version 1.2) says prompts, replies, attachments, generated images and audio, chat history and models stay on the device, that recordings are processed in memory and not uploaded, and that there is no LLM Studio Go account and no separate analytics SDK. Model downloads contact Hugging Face, which receives ordinary connection data such as your IP address; with web search switched on, the query goes to DuckDuckGo and pages are read from the sites visited. The optional local API uses plain HTTP, so the policy advises a trusted network only.

The ads are the main privacy trade-off. Per the policy, Google AdMob shows them after a Google consent prompt and, where applicable, Apple's App Tracking Transparency prompt; Google and its partners may process your IP address, approximate location, device or advertising identifiers and ad interactions, while chats, images, recordings and model files are not sent to Google for advertising. The App Store privacy labels, which the developer provides and Apple does not verify, list Identifiers (Device ID) as data used to track you; Coarse Location, Identifiers and Usage Data (product interaction, advertising data) as data linked to you for third-party advertising; and diagnostics data not linked to you for advertising and analytics. Declining tracking does not disable the local AI features.

📌Note: These are the developer's declarations, not audit results, and no source code is published. PromptQuorum has not inspected the app's network traffic or its ad SDK.

What We Could Not Verify

  • Hands-on behavior. PromptQuorum did not run the app, so answer quality, speed, battery use, the quality of music and voice output and the memory each model needs are unknown.
  • What the purchase changes. The sources do not say whether the ad SDK stops collecting data after Remove Ads Forever, and they do not reconcile the policy's "no separate analytics SDK" with labels that list analytics diagnostics.
  • Source code and license. No repository or license text was found, so the privacy and offline statements cannot be checked against code.
  • Track record. The app is one week old and showed no App Store ratings when checked.
  • Model catalog. The listing does not name the models offered, the engines used or which models count as supported or compatible for vision, images and audio.

Competitors and Alternatives

Locally AI

Platforms:
iPhone, iPad, Mac
Price / license:
Free / closed source
Key difference:
Runs open models on the device through MLX
Articles about Locally AI (5)

Also mentioned in:

LLM Hub

Platforms:
Android, iOS
Price / license:
Freemium / source-available
Key difference:
Published source under a non-commercial license
Articles about LLM Hub (1)

Also mentioned in:

Competitor details change often; confirm each app's current price, license and platforms on its own listing.

Compare in directory

Frequently Asked Questions

Which languages does the app support?

The App Store lists 12 interface languages: Arabic, English, French, German, Hindi, Japanese, Korean, Portuguese, Russian, Chinese, Spanish and Ukrainian.

What do I need to run it, and how large is the download?

iOS or iPadOS 18.0 or later; the app itself is about 129 MB, and models come on top. The listing calls audio models large downloads and says availability and speed depend on the device, free storage and the model.

Does it need an account?

No LLM Studio Go account is needed, per the policy; your Apple Account is used only by the App Store, for example to restore the purchase.

Is it suitable for children?

The policy calls it a general-purpose productivity tool that is not directed to children; because it shows Google ads, parents may want to check the ad privacy options in its Settings.

Verdict

Per its listing, LLM Studio Go gives iPhone and iPad owners a free way to run chat, image, voice, music and speech models locally, with optional web search and a local API for your own network. Against that, no source code is published, it shows Google ads whose App Store privacy labels list tracking identifiers, was one week old with no ratings when checked and has not been tested by PromptQuorum. It suits people who want one free app for several kinds of local generation and accept the ads; readers who want published code can compare PocketPal AI, and those who prefer a paid app can compare Private LLM.

Sources

← Back to Power Local LLM

MCP