# Open Relay - Open WebUI Client

> Native iOS client for Open WebUI to chat with AI models on your server. (iPhone/iPad app by Abhishek Bhardwaj.)

- Source: https://appshunter.io/ios/app/open-relay-open-webui-client/id6759630325 (this page in markdown: same URL + `.md`)
- Developer: [Abhishek Bhardwaj](https://appshunter.io/developer/1777945036)
- Category: Productivity, Entertainment
- Price: $9.99
- Rating: 5.00/5 from 25 App Store ratings · 24 written reviews indexed
- Age rating: 4+
- Requires: iOS 18.1 · 79 MB
- Languages: American English, Arabic, Azerbaijani, Basque, Bangla, Bosnian, Bulgarian, Catalan, Cebuano, Croatian and 44 more
- Released: 2026-03-11
- Data updated: 2026-08-28
- Monetization: paid
- User reviews in markdown: https://appshunter.io/ios/app/open-relay-open-webui-client/id6759630325/reviews.md

## What is Open Relay?

Your AI. Your server. Your pocket.

Open Relay is a beautiful, 100% native iOS & iPadOS client for Open WebUI. Chat with any model on your server, right from your iPhone or iPad, with a fast, polished experience that feels native. 

— STREAMING CHAT WITH RICH MARKDOWN —
Watch responses flow in word by word with full markdown rendering — syntax-highlighted code blocks, tables, math equations, Mermaid diagrams, inline SVG images, and more. Everything renders beautifully in real time.

— VOICE CALLS WITH AI —
Call your AI like you'd call a friend. Open Relay uses Apple's CallKit so it looks and feels like a real phone call — complete with an animated orb that reacts to your voice and the AI's response.

— ADMIN CONSOLE —
Take full control of your WebUI admin settings directly from the app instead of visiting the site. Directly add new integrations, mcp servers, or modify server settings quickly through the app. 

— TERMINAL INTEGRATION —
Give your AI access to a real terminal. Enable command execution from chat, then swipe from the right edge to open a file browser with directory navigation, file upload, preview, and a mini terminal.

— SWITCH MODELS ON THE FLY —
Type @ to mention any model and switch mid-conversation. A persistent chip in the composer shows which model is active.

— PROMPT LIBRARY & SKILLS —
Type / to search your prompt library with slash commands, or $ to browse and apply your Open WebUI skills — all without leaving the composer.

— KNOWLEDGE BASES (RAG) —
Type # to search and attach your knowledge collections, folders, and files. Works exactly like the web UI.

— ON-DEVICE VOICE, COMPLETELY PRIVATE —
Built-in neural text-to-speech (Marvis) and speech-to-text run entirely on your device via Apple's MLX framework. Download once, then no data ever leaves your phone. Server-based and Apple system voices also supported.

— TOOLS & WORKSPACE —
Toggle server-side tools on or off per conversation. Tool calls render inline with collapsible details. Manage models, knowledge bases, prompts, skills, and tools directly from the app.

— CHANNELS —
Collaborative chat rooms where multiple users and AI models interact — organized by Direct Messages, Groups, and Channels.

— AUTOMATIONS —
Automations let you schedule prompts to run automatically at recurring times.

— TASK MANAGEMENT —
Task Management gives agentic models a structured way to plan and track multi-step work in a chat. 

— CALENDAR —
Schedule, track, and manage events — with AI that can plan for you.

— ORGANIZE EVERYTHING —
Drag and drop chats into folders. Pin important conversations. Search across everything. Configure folders with system prompts, default models, and attached knowledge. Archive and restore in bulk.

— RICH ATTACHMENTS —
Attach files, photos, or paste images into chat. The Share Extension sends content from any app straight into a conversation.

— DEEP THEMING —
Accent color presets or a custom color wheel. Pure black OLED mode with tinted surfaces. Every change previews live.

— MULTI-SERVER/ACCOUNT SUPPORT —
Save multiple Open WebUI connections and switch instantly. Full auth support — username/password, LDAP, SSO, and auth proxies like Authelia and Authentik.

— HOME SCREEN WIDGETS & SHORTCUTS —
Start a chat from a widget or your Action Button. Background notifications alert you when a generation finishes while you're in another app.

 — ACCESSIBILITY — 
Independently adjust message text size, conversation title size, and UI element scaling

Requires a running Open WebUI server instance.

Your models. Your rules. Beautifully native.

## Key features

- Native iOS & iPadOS client
- Streaming chat with rich markdown
- Voice calls with AI
- Admin console for server settings
- Terminal integration with file browser
- On-device voice processing
- Prompt library and skills integration

## Chart rankings

- #58 in Top Paid · Productivity (US App Store)

## Recent user reviews (10 of 24)

All indexed reviews: https://appshunter.io/ios/app/open-relay-open-webui-client/id6759630325/reviews.md

### 2/5 — Good idea, but very buggy

*2026-08-22*

When I bought the app originally I was quite happy with it. The field is just developing, bugs and other issues here and there are expected. But none of the last updates did for me what they suggested they would. Streaming still breaks, thinking / reasoning is wrongly displayed or breaks the stream entirely, and entire conversations vanish. Looks like as if this app needs a proper rewrite in its streaming engine.

**Developer response:** Hello, sorry to hear you are experiencing such issues. I am actively working with individuals through github fixing this as it does not happen to the majority but specific models. I myself is unable to replicate the issue hence why i have to depend on the community to help fix this for such models. Regarding your suggestion about a rewrite, its already in works and i have started breaking it down completely and hopefully building a more robust system that can cover all edge cases. Lot of the early development was focused on performance which has been obtained so reworking the streaming engine would make sense. Thank you for all the support and i look forward to getting it all fixed! Please follow github for further progress as there is an ongoing issue thread related to this.

### 5/5 — Hands Down Best Open Web Ui

*2026-08-15*

The Open Web Ui project needs to adopt this app and developer as its official mobile app. It’s always solid, the Developer is active. Takes feedback, proud supporter. Thanks mate keep up the great work.

### 4/5 — Great Overall! Server switching bugs

*2026-08-14*

I love this app, and the only reason I remove one star is I am regularly switching between a local connection and a tailscale connection. It doesn’t seem to handle switching servers regularly very well. I will switch to the server I know I can reach, it hangs trying to connect, then I must switch back and forth before restarting the app. It would be nice if you could exit the switch server page without switching servers using a (X) in the corner (since it requires you to switch a server).

**Developer response:** Hello, I am glad you are enjoying the app and thank you for raising that concern. I have pushed an update to improve the multi-server experience. Feel free to open an issue on github if it does not meet your requirements and I'd be happy to help.

### 5/5 — The best

*2026-08-12*

This is under very active development. It is already great, and just getting better.

### 5/5 — Excellent Experience vs Mobile Web

*2026-07-27*

I started using this app on its early versions; and initially it didn’t have every feature I need (I use OWUI heavily; for everything) - but each time I submitted a bug report of feature request to the GitHub, a few days later it was done.

Now some months later I can confidently say I am coming to this app instead of the mobile web version every time. The developer has proven this is a long term, maintained and enhanced project, and I can recommend it to you without hesitation.

### 4/5 — Top OpenWebUI Clien App

*2026-07-23*

War auf der Suche nach einer guten alternative zur offiziellen PWA App. Bisher super zufrieden und Entwicklung noch in Arbeit. Auch OpenWebUI Einstellungen sind in einer modernen Oberfläche erreichbar.

### 5/5 — Open Web Ui - Must Have

*2026-07-16*

I would vote this to become the official iOS for Open Web Ui. That’s how well maintained this app is. For now, the development of the app is second to none. 5/5

### 5/5 — Better UI than WebUI in Browser

*2026-07-12*

Works very well and bugs are fixed quickly.

### 5/5 — Incredible App

*2026-06-27*

Really made me fall in love with connecting to my laptop, it's so fast and effortless but a few suggestions (I don't know how technically possible these are so they're just things that I would love to have if possible):

1. Automatic message queuing doesn't auto send the message even when I check the setting to do so, I have to manually press the up arrow on the queued message to send it - so if it can auto send as soon as the last prompt is done that would be great.

2. Sometimes it isn't clear if the server is already processing a message so it looks like it's taking an extremely long time when in reality behind the scenes the computer is processing 2 previous prompts and the third is in queue. I'd love to either see how many prompts are actively being processed and either stop all or some of them. My work around is to load a new model which flushes out all the queued prompts and start over.

3. Ability to load 2 models at once - not sure if JIT won't allow this but I'd love the ability for you to try and if for whatever reason there isn't enough memory it just tells you or fails to load.

4. More details on token usage like total input/total output, maybe a per model tracker. Maybe for fun a built in calculator you put in $X input and output costs and it estimates "you saved $XY running local!" So you can compare it.

5. (No idea if this is possible), a model tracker for how many tokens / max you have. So if a model can take 260k it shows you 80k/260k in a little tracker instead of pressing the little i.

### 5/5 — Worth it

*2026-05-26*

I was hesitant at first - do I really need a rather expensive app for an application that has a perfectly fine web UI? Yes, in fact I do.

## Frequently asked questions about Open Relay

### What is Open Relay?

Open Relay is a 100% native iOS and iPadOS client for Open WebUI. It allows you to chat with any AI model hosted on your server directly from your iPhone or iPad with a fast, polished experience.

### Does Open Relay support voice calls with AI?

Yes, Open Relay supports voice calls with AI using Apple's CallKit. It integrates seamlessly to feel like a real phone call, complete with an animated orb that reacts to your voice and the AI's response.

### Can I manage my Open WebUI server settings from Open Relay?

Yes, Open Relay includes an Admin Console that lets you take full control of your WebUI admin settings directly from the app. You can add integrations, modify server settings, and manage other configurations without visiting the website.

### How does Open Relay handle AI responses?

Open Relay offers streaming chat with full markdown rendering. This means responses flow in word by word, and it beautifully displays syntax-highlighted code blocks, tables, math equations, Mermaid diagrams, and inline SVG images in real time.

### Is Open Relay available on Android devices?

No, Open Relay is exclusively a native iOS and iPadOS client. It is designed to work on iPhone and iPad devices.

### What is the price of Open Relay?

Open Relay is a paid application, priced at $9.99. This one-time purchase grants you access to all its features.

### Does Open Relay have ads?

No, Open Relay does not contain advertisements. As a paid application, it provides an ad-free experience for its users.

### How often is Open Relay updated?

The latest version of Open Relay is 4.9, which was last updated on June 2, 2026. Updates are released regularly to enhance features and performance.

## Version history (last 5 releases)

### 5.4.2 — 2026-08-25

IMPROVEMENTS
• Responses now show a notification when interrupted mid-stream (e.g. if the app was backgrounded too long), so you always know to tap back in and retry.
• Channel message notifications no longer stack up as duplicates from the same person — each sender gets one banner that updates in place.
• App badge now increments when responses or channel messages arrive, and clears when you tap a notification.

BUG FIXES
• Fixed tool call responses being dropped when a model executes multiple tools concurrently.
• Fixed a bug where content could appear doubled in responses that use the structured output format.
• Fixed a bug where Socket.IO connection "user-join" handshake could silently leak memory if the app was backgrounded during the 5-second acknowledgement window.
• Fixed a race condition in the Socket.IO polling transport where rapid reconnects could spawn two simultaneous polling loops, causing duplicate or dropped events.
• Fixed the streaming URL session not being reset when switching servers, which could send requests to the wrong server after a switch.
• Fixed a bug where the same AI response could fire two separate "response ready" notification banners (one from background polling and one from foreground recovery).

### 5.4.1 — 2026-08-24

BUG FIXES
• Fixed admin users unable to edit models in Admin Console
• Fixed starred tool quick-action pills disappearing on app launch and new chat

### 5.4 — 2026-08-22

WHAT'S NEW
• Added Controls panel (tap the slider icon in any chat) showing attached files, knowledge bases, reference chats, system prompt override, and tool/function valve settings

IMPROVEMENTS
• Improved server switching: the "Switch Server" screen now has an X button to dismiss without switching, server rows show a reachability check before committing a switch, and the connection-lost overlay presents the switcher as a dismissible sheet.

BUG FIXES
• Fixed starred tool buttons persisting in the quick-actions row after a tool is deleted or removed from the server.
• Fixed hidden models not appearing in the workspace model editor's base model picker — admins can now select any enabled model (including hidden ones) as a base model when creating or editing workspace models, matching Open WebUI web behavior.

### 5.3.9 — 2026-08-21

IMPROVEMENTS
• Admin Models screen now shows workspace custom models (e.g. models you've created from a base) alongside base models.

BUG FIXES
• Fixed second (and subsequent) reasoning/thinking blocks not appearing during streaming.
• Fixed admin model list only loading the first 30 models when more than 30 exist.
• Fixed all users appearing to have edit access to models they don't own; edit button is now greyed out for read-only models.
• Fixed "Hide Model" / "Show Model" failing for Ollama and other base models that don't yet have a workspace record.
• Fixed admin model settings (built-in tools toggles) always resetting to default because the server key names didn't match what the app expected.
• Fixed model public/private badge always showing "PRIVATE" due to access grants being checked in the wrong place.

### 5.3.8 — 2026-08-20

BUG FIXES
• Fixed follow-up suggestions disappearing ~2 seconds after appearing due to server sync overwriting them with an empty array.
• Fixed follow-up suggestions not persisting when editing a message or switching between message versions.
• Fixed follow-up suggestions never appearing after regenerating a response.
• Fixed reasoning blocks displaying as raw HTML instead of rendering natively during streaming.

## More apps by Abhishek Bhardwaj

- [Augmented Imagination](https://appshunter.io/ios/app/augmented-imagination/id6737791200)
- [Picklin' Pals](https://appshunter.io/ios/app/picklin-pals/id6744089595)

All apps by Abhishek Bhardwaj: https://appshunter.io/developer/1777945036

## Related topics

[open webui](https://appshunter.io/ios/topics/open-webui)

---

*Data collected daily from the US App Store and indexed by [AppsHunter](https://appshunter.io/). User reviews are verbatim App Store reviews. Ratings, prices and chart positions refresh continuously; this snapshot is from 2026-08-28.*
