Skip to main content
PromptQuorumBuilt for humans. Structured for AI.
Home/Power Local LLM/Tina Review: iPhone Client for Ollama, llama.cpp and ComfyUI
Mobile & Edge LLMs

Tina Review: iPhone Client for Ollama, llama.cpp and ComfyUI

·7 min read·By Hans Kuepper · Founder of PromptQuorum · Discovery engine for open-weight & open-source AI
Read in:
🇺🇸en🇩🇪de🇫🇷fr🇯🇵ja🇨🇳zh🇪🇸es🇧🇷pt🇸🇦ar🇰🇷ko

Tina is a freemium, closed-source Apple app that connects to models you host yourself (Ollama, llama.cpp), generates images through ComfyUI, and can also run models on the device. Per its App Store listing it needs iOS, macOS or visionOS 26.0 or later, costs nothing to download, and sells Tina Pro at $1.99 and a $12.99 yearly option as in-app purchases. This review covers version 1.3.8 and is based on the public listing, not hands-on testing.

Tina is an iPhone and Mac app that chats with models on your own server, in your homelab, or on the device itself, and also drives ComfyUI image generation. This review covers what its App Store listing documents, what is unclear, and how Tina compares to other mobile clients for self-hosted models.

Tina Review: iPhone Client for Ollama, llama.cpp and ComfyUI

Key Takeaways

  • Version reviewed: 1.3.8, as shown on the App Store listing.
  • Price: free to download; Tina Pro $1.99 and a $12.99 yearly option are listed as in-app purchases.
  • Connections: Ollama, llama.cpp, on-device models and ComfyUI, plus Home Assistant, Firecrawl, Proxmox and RunPod integrations.
  • Platform: iPhone, Mac with Apple silicon, and Apple Vision Pro, all on version 26.0 or later.
  • Openness: closed source; no public repository or product site was found.

📍 In One Sentence

Tina, by developer Wes Wickwire, is a freemium Apple-platform client that chats with your own Ollama or llama.cpp server, generates images through ComfyUI, and also runs on-device models.

💬 In Plain Terms

Think of it as a remote control for the AI you host at home, with a small on-device option added: the heavy models stay on your server, and the app on your phone is the interface.

Get Tina

Tina is distributed through the Apple App Store. The download is 56.3 MB and the listing is rated 4+.

This review is a companion to PromptQuorum's Local LLM Software Directory, which lists Tina alongside other mobile and desktop local AI tools.

Channel
Get It
Apple App StoreTina - Private Local AI on the App Store
Privacy policyTina privacy policy

No dedicated product website or source repository was found for Tina. The App Store listing is the main public source.

Tina at a Glance

Attribute
Tina
PlatformiOS, macOS, visionOS (26.0+)
PriceFree, in-app purchases
LicenseClosed source
Runs fully offlineOn-device models only; server mode needs a link
Built-in engineYes, on-device models are listed
Works with OllamaYes
Image generationYes, via ComfyUI with a node editor
Voice input / outputYes, voice and text-to-speech

Attributes follow the mobile-chat comparison used in the Local LLM Software Directory. They come from the App Store listing alone, which does not cover every feature.

What Tina Is

Tina is a client first and an on-device runner second. The listing's pitch is "connect to local models: your own server, on your device, or your homelab", with Ollama, llama.cpp and on-device models named as sources.

It is published by Wes Wickwire, an individual developer, and is categorized as a Utility. No inference engine for the on-device models is named, so this review does not state which runtime it uses.

This review draws on the App Store listing only. It does not include hands-on testing, so speed, stability and answer quality are not rated here.

How to Get Started

Setup is an App Store install, then pointing the app at a server or a model. The listing does not document the first-run flow, so the steps below follow what it states.

  1. 1
    Check your OS version
    Why it matters: The listing requires iOS, macOS or visionOS 26.0 or later, and on Mac an Apple M1 chip or newer.
  2. 2
    Install Tina
    Why it matters: Get it from the [App Store](https://apps.apple.com/us/app/tina-private-local-ai/id6751924571); the download is 56.3 MB.
  3. 3
    Connect a model source
    Why it matters: Add your Ollama or llama.cpp server address, or choose an on-device model, depending on where you want inference to run.
  4. 4
    Optional: add ComfyUI and integrations
    Why it matters: Link a ComfyUI instance for image generation, and Home Assistant, Firecrawl, Proxmox or RunPod if you use them.

Features and Integrations

Tina's selling point is reach: one app for chat, images and homelab tools. Everything below comes from the App Store listing.

  • Chat with LLMs. Talk to models served by Ollama or llama.cpp, or to models on the device.
  • ComfyUI image generation. The listing describes full ComfyUI support with a node editor inside the app.
  • Voice. Voice and text-to-speech support are listed.
  • Homelab integrations. Home Assistant, Firecrawl, Proxmox and RunPod are named; the listing gives no detail on what each integration can do.
  • Multiple Apple platforms. One listing covers iPhone, Mac and Apple Vision Pro.

Device Requirements

Tina requires version 26.0 or later of iOS, macOS or visionOS, and an Apple M1 chip or newer on Mac. The listing gives no RAM figure.

Where inference runs decides the real hardware need. With a remote Ollama or llama.cpp server the phone only needs to show the chat; with on-device models, the iPhone's memory is the limit.

Pricing and Privacy

Tina is free to download, with Tina Pro at $1.99 and a yearly option at $12.99 listed as in-app purchases. The listing does not say which features Pro unlocks, so check the paywall in the app before relying on a feature.

Apple's privacy section says Tina collects usage data and diagnostics (crash and performance data) that are not linked to your identity. Apple notes these labels are self-reported by the developer and not verified.

Chats sent to your own server travel to that server, so your privacy there depends on how you host and secure it. The app is closed source and this review did not inspect network traffic.

Trade-Offs: Benefits vs. Limitations

Bring your own server

What it means in real use:
Run large models on a home machine and chat from the phone.
Limitation / caveat:
You must set up and expose Ollama or llama.cpp yourself.

ComfyUI built in

What it means in real use:
Chat and image generation live in one app, including a node editor.
Limitation / caveat:
It needs your own ComfyUI instance; the listing does not name supported workflows.

On-device option

What it means in real use:
Some chats work without any server.
Limitation / caveat:
No model list, size or runtime is published for the on-device mode.

Low entry price

What it means in real use:
Free to try; Pro is $1.99 as listed.
Limitation / caveat:
The listing does not explain what Pro adds, or whether $1.99 is one-time.

Apple-wide support

What it means in real use:
One purchase path covers iPhone, Mac and Vision Pro.
Limitation / caveat:
Everything requires version 26; there is no Android or Windows app.

Tina vs. Alternatives

Tina

Platforms:
iOS, Mac, Vision Pro
Price and license:
Freemium, closed source
Model source:
Ollama, llama.cpp, on-device
Key difference:
Adds ComfyUI images and homelab integrations; not open source

Enchanted

Platforms:
iOS, Mac
Price and license:
Free, Apache 2.0
Model source:
Ollama
Key difference:
Open-source Ollama client with no built-in engine of its own
Articles about Enchanted (4)

Also mentioned in:

+5 more not shown

Maid

Platforms:
Android, iOS
Price and license:
Free, MIT
Model source:
On-device, Ollama, OpenAI API
Key difference:
Open source and on Android too; ComfyUI not listed
Articles about Maid (5)

Also mentioned in:

AnythingLLM Mobile

Platforms:
iOS, Android
Price and license:
Free, MIT
Model source:
Your AnythingLLM workspace
Key difference:
Remote access to one AnythingLLM instance rather than raw model servers
Articles about AnythingLLM Mobile (2)

Also mentioned in:

Platform, price, and feature details for third-party apps change frequently. Verify current specifics on each app's own listing before deciding.

Compare in directory

Who Should Use Tina

  • Homelab owners who run Ollama or llama.cpp. Tina is built around connecting to those servers from an Apple device.
  • ComfyUI users who want mobile access. The node editor and generation controls are part of the app.
  • Users who want a server and an on-device fallback in one app. Both modes are listed.

Who Should Not Use Tina

  • Anyone who needs open source. The code is not public; Enchanted is open source for Ollama.
  • Android or Windows users. Only Apple platforms are listed; Maid covers Android.
  • People below iOS or macOS 26. The listing requires version 26.0 or later.
  • Anyone without a server who wants a model catalogue. The on-device mode has no published model list; see Locally AI.

Frequently Asked Questions

Is Tina free?

It is free to download with in-app purchases: Tina Pro at $1.99 and a yearly option at $12.99, as shown on the App Store listing.

Who makes Tina?

Wes Wickwire, an individual developer listed on the App Store. A GitHub profile of the same name hosts open-source Swift libraries, but no Tina repository.

Do I need my own server to use Tina?

No, the listing also names on-device models. A server is only needed for Ollama, llama.cpp and ComfyUI features.

Does Tina work on Android?

No. The listing covers iPhone, Mac with Apple silicon and Apple Vision Pro only.

Which integrations does Tina list?

Home Assistant, Firecrawl, Proxmox and RunPod. The listing gives no detail on how each one works.

Verdict

Tina is a good fit on paper for someone who already hosts models at home and wants a single native Apple app for chat, ComfyUI images and a few homelab hooks, with an on-device fallback.

The caveats are transparency and age: it is closed source, the on-device engine and Pro features are undocumented, and the only public source is the App Store listing, which showed three ratings at the time of review.

If you run Ollama and want open source, start with Enchanted. If you want ComfyUI and homelab integrations in the same app, try Tina's free tier first.

Sources

← Back to Power Local LLM