What It Is
AI Tool Integrations describe how PANTA OS combines its own model layer with specialized AI providers for media generation. These providers handle modalities that the general model layer does not produce natively: voice generation, avatar generation, and video composition. Inside PANTA OS, they appear as steps within workflows and apps, not as separate tools selectable in chat. The platform ships with a set of these AI providers already wired in. Additional ones can be added as needed, set up by the PANTA OS team rather than self served by customers. This page is the strategic overview of the AI provider class. For the day to day handling of standard tool integrations like Microsoft 365, Google, Notion, GitHub, and Microsoft Teams in chats and assistants, see Integrations.Why It Matters
The general model layer is good at text, code, and analysis. For voice, avatars, and video, specialized AI providers produce better and more controllable results. Bringing these providers into PANTA OS at the workflow level avoids three problems.No tool sprawl
Users do not maintain separate accounts and separate billing with each AI provider. The providers are wired into the platform; users just run the workflow.
Modalities chat does not cover
Voice generation, avatar videos, and full video compositions are not chat outputs. Routing them through workflows is the right shape for the work.
Repeatable media production
A workflow that combines a script step, a voice step, and an avatar step produces consistent media output every time it runs.
Governance stays in one place
Spend, usage analytics, and access controls live inside PANTA OS, not spread across each individual provider’s dashboard.
Two Integration Classes
PANTA OS distinguishes between two kinds of integrations. Both feed the same goal but live in different parts of the product.Standard tool integrations
Productivity systems used inside chats and assistants for everyday work. The standard set covers Microsoft 365, Google, Notion, GitHub, and Microsoft Teams. See Integrations for handling.
AI tool integrations
Specialized AI providers used as steps inside workflows and apps. This page describes that class.
Where AI Tool Integrations Show Up
AI tool integrations are not exposed as a list of services in the user interface. They sit inside workflows and apps as the engine that drives specific step types.Voice steps
A workflow step that converts a script into spoken audio. The voice generation provider is selected and configured by the workflow builder; the user just runs the workflow.
Avatar steps
A step that animates a presenter avatar against a script and voice track. Output is a video clip with a synchronized avatar.
Video composition steps
A step that assembles avatar clips, voice, intros, outros, and overlays into a finished video. The output lands in the Library when the workflow completes.
Image generation
Image generation is available natively in chat without a separate AI tool integration. AI tool integrations specifically cover voice, avatar, and video, where multiple steps need to coordinate.
Adding More AI Tools
The set of integrated AI providers is curated. Additional providers are added when a customer use case requires them, configured by the PANTA OS team rather than connected by the customer.What customers can request
A specific provider, a modality not yet covered, or a step type that combines multiple providers into one experience.
What customers do not configure
Authentication, billing, and the technical wiring of the provider. The PANTA OS team handles the connection so the result is a clean step inside workflows.
Tips and best practices
- For voice, avatar, and video work, build a workflow. Free form chat is not the right shape for media production.
- Keep media outputs in the Library for reuse. The same script and avatar combination produces consistent output across runs.
- Watch the Inbox for media generation completions. Some workflows that take longer to produce videos, images, or audio output may be delivered via email.
- For an AI provider that PANTA OS does not integrate yet, talk to your account contact. New providers are added on demand.
Help Center
Where do I pick a voice provider or video provider in the chat
Where do I pick a voice provider or video provider in the chat
You do not. AI tool integrations are not selectable in chat. They live inside workflows and apps as step types. You will work with the PANTA OS team to confirm the details of the audio output (language, voice ,etc).
Why is image generation handled differently from video
Why is image generation handled differently from video
Image generation works natively in chat without picking a separate model. Video composition spans multiple steps (script, voice, avatar, assembly) that have to run together, which is why it lives in workflows as a sequence of AI tool integration steps.
Can my organization add a new AI provider
Can my organization add a new AI provider
On request. The PANTA OS team configures new AI provider integrations as part of customer onboarding or expansion. Customers do not self serve these connections.
How is spend on AI tool integrations controlled
How is spend on AI tool integrations controlled
Through the same Token limits in the Admin Panel that govern model spend. The platform tracks usage across all integrated providers; administrators set limits per model and a workspace plan budget.
What is the difference between this page and the Integrations page
What is the difference between this page and the Integrations page
This page covers AI providers used inside workflows and apps for media generation (voice, avatar, video). The Integrations page covers standard productivity tools (Microsoft 365, Google, Notion, GitHub, Microsoft Teams) used inside chats and assistants. Two classes of integrations, two pages.