Run aiXplain Agents from Your Phone
We have just made agent collaboration radically simpler.
Starting today, any team on aiXplain can deploy agents and interact with them directly inside the aiXplain Chat app for iOS. No frontend coding. No integration steps. Just build, deploy, switch to your team, and chat.
It’s a faster way to put AI to work across your organization, giving your team an interactive interface to any agent you’ve built, complete with memory, file exchange, and real-time tool usage. And it’s all running in a secure, usage-based model that scales with you.
What’s New in the aiXplain Chat App
Now, teams can share and run agents directly from the aiXplain Chat app. Just switch to your team workspace using the team switcher, and all deployed agents for that team are instantly accessible.
Agents run inside a familiar chat interface:
- Upload and download files
- Run tools and pipelines
- Maintain context over multi-turn conversations
- See outputs as they happen, no custom UI needed
Best of all, your agents feel like apps, but without the overhead. They’re reusable, interactive, and composable tools your team can invoke at any time.
Built for Developers, but Accessible to Everyone
Whether you’re using the SDK or Agent Builder on macOS, you can create and deploy an agent in minutes.
SDK: Check out this Colab for how to build an AI agent designed to retrieve and analyze content from Hacker News
Agent Builder: Agent Builder on macOS lets you describe your agent, drag in tools from aiXplain’s rich marketplace, and deploy with one click. Orchestration, hosting, and scaling are all handled by the aiXplain Agentic OS. No infrastructure required.
The video below shows a deployed Hacker News Agent in Agent Builder and running it in the Chat app.
Simple Economics That Scale With Usage
You don’t pay to host the agent. You don’t need to manage the compute. Costs only apply when your agent is actually used.
When a teammate runs your agent via the Chat app:
- It draws from the team’s credit wallet
- No charges are made to personal accounts
- All usage is tracked transparently in your dashboard
This makes agent access scalable, permission-aware, and cost-predictable. It’s ideal for organizations that want to make AI accessible across roles and departments without spinning up custom tools or billing overhead.
What Teams Are Already Building with AI Agents Today
We’re seeing early use cases across teams and industries:
- Internal support agents that answer product and process questions
- Meeting summarizers that handle file uploads and language conversion
- Sales agents that draft outreach using your own pitchbook
- Technical assistants that run pipelines or benchmark models on demand
- Research assistants that analyze text and PDF documents
Now, each of these agents becomes immediately usable through the Chat app, enabling real-time interaction without custom dev work.
How to Get Started With the aiXplain Chat App
This feature is currently in beta for teams on Builder, Team, and Enterprise plans.
To try it:
- Create and deploy an agent using the SDK or Agent Builder
- Open aiXplain Chat app on iOS
- Use the team switcher to enter your team’s workspace
- Start chatting with any deployed agent
In addition, you can still access all public aiXplain agents and models from within the app, and even swap LLMs mid-conversation to compare responses or optimize performance.
Current Limitations
While, this is just the beginning. Today’s release does not yet support:
- External data integrations directly from the Chat app
- Persistent memory beyond session scope
- Role-based access control per agent (coming soon)
However, we’re actively building toward more granular permissions, agent visibility settings, and the ability to embed agents in custom apps or workflows.
Bringing Agent Collaboration to Life With the aiXplain Chat App
We believe AI agents should feel less like distant services and more like team members. This update brings us closer to that vision:
- Build an agent once
- Share it with your team instantly
- Run it anywhere your team works
Ultimately, this is part of our broader mission to make multi-agent systems usable, scalable, and production-ready for real-world teams.
The future of AI isn’t just smarter models. It’s a smarter collaboration.
We have cookies!