MlyticsAGIC
Contact sales

Platform

Model APIs
and global delivery

Use agents and model APIs on top of Mlytics compute, integrated model providers and global delivery.

View API pricingDiscuss your architecture

Model APIs

Language, video,
voice and image APIs

Call models directly or let the agent coordinate requests. Supported routes use Mlytics compute or integrated model providers, with usage metered by route.

Video · catalogue preview

Video API

Generate shots from prompts and approved imagery. Formats and duration depend on the selected model route.

Metered per generated video-second
Illustrative request 
{
  "capability": "video",
  "input": "A quiet morning. A fresh start.",
  "output": {
    "duration_seconds": 15,
    "aspect_ratio": "9:16"
  }
}
Example payload · final schema follows the API catalogue

Model, format, language and voice options depend on the route. Ask about Taiwan Mandarin, Taiwanese Hokkien and Hakka deployments.

Model access, with clear boundaries

Choose where your models run

Model hosting, usage meters and deployment requirements follow the selected service.

Mlytics compute

Supported self-hosted models

Mlytics-owned GPUs provide core compute for supported open-weight models. Model selection and job scheduling belong to the AI orchestration layer.

Integrated model providers

Provider-served model routes

Access additional capabilities through authorised provider services. Their model, processing location and usage terms remain specific to that route.

Cloud-neutral integration means connecting to your existing systems. Data routes, retention and model rights follow the selected deployment and agreement.

Compute and delivery

Model execution
and global delivery

Run supported models on Mlytics compute or through integrated providers. Decisive Engine routes approved assets across Multi-CDN for regional delivery.

AI compute and model routes

Mlytics-owned GPUs serve supported self-hosted models. Integrated providers serve their own model routes. AI orchestration coordinates model requests, generation jobs and asset versions.

Self-hosted modelsProvider model routesAI job orchestration
Connect your existing cloudReduce single-CDN dependenceItemised generation and delivery usage

Your creative, your destinations

From approved asset
to audience experience

Publish approved assets to your websites, applications and campaign destinations. Connect eligible creative with Cortex AI Mode where campaign inventory is available.

Three clear responsibilities

  • AI orchestrationCoordinates models, generation jobs and asset versions.
  • Decisive Engine & Multi-CDNCoordinate content delivery across available CDN routes.
  • Cortex AI ModeConnects relevant content experiences with market context and brand partnerships.
Explore market activation →

Questions

Before production

Does every model run on Mlytics-owned GPUs?

Supported self-hosted models can run on Mlytics compute. Provider model routes run through their authorised services. Confirm the model, hosting and data path before production.

Does Decisive Engine schedule the GPUs?

AI orchestration coordinates model execution and generation jobs. Decisive Engine handles delivery-routing decisions across CDN providers.

Does global delivery include media placement?

Global delivery distributes approved assets to the configured destinations. AI Mode media availability, target-market coverage and campaign budget are agreed separately.

Discuss your model and delivery requirements

Discuss model access, delivery requirements and the markets you want to serve.

Contact sales