M

MasterA AI GPU Inference

Published byAnestis Demeroutis
9/9/2026
Agent payments

Endpoints

10

Image Generation
7

Video Generation
3

What is the MasterA AI GPU Inference API?

High-speed AI image & video generation on our own NVIDIA RTX GPUs (RTX 5090, RTX 5080 & RTX 4070 Ti Super) — Flux.1-Schnell, SDXL-Turbo, and LTX-Video. One API key, EU data locality, solar-powered 24/7.

MasterA AI GPU Inference API is available on the JoJ API marketplace. Subscribe with a single JoJ API key, call it through one gateway, and pay from one wallet — no separate account or contract with the provider.

Pricing & plans About this API FAQ Discussions More APIs by @anestis-demeroutis OpenAPI spec MCP server

MasterA AI GPU Inference API endpoints

  • POST/v1/ltx/generate

    ltx-text-to-video

    Generate a short video (~4s clip) from a text prompt using LTXV-13B-0.9.8-distilled. Returns a task_id immediately — poll GET /v1/tasks/{task_id} for the result.

  • GET/v1/tasks/{task_id}

    get-task-status

    Poll generation progress. All generation endpoints are asynchronous and return immediately with a task_id — use this endpoint to check status and retrieve the final result. Status values: processing (queued), generating (running on GPU), completed (ready, with output_url), error (failed).

  • GET/v1/health

    video-health

    GPU, service and loaded-model health for the video server.

  • GET/v1/files/{filename}

    get-file

    Download a generated image (PNG) or video (MP4). The filename is taken from the output_url returned by GET /v1/tasks/{task_id} when status=completed.

  • POST/v1/flux/generate

    flux-text-to-image

    Generate a high-quality image from a text prompt using Flux.1-Schnell. Returns a task_id immediately — poll GET /v1/tasks/{task_id} for the result.

  • POST/v1/sdxl/generate

    sdxl-text-to-image

    Generate an image in ~2 seconds using SDXL-Turbo. Best for high-volume, real-time, and prototyping use cases. Returns a task_id immediately — poll GET /v1/tasks/{task_id} for the result.

  • GET/v1/status

    model-status

    Returns which model is currently loaded on the GPU and its idle time.

  • GET/v1/video/tasks/{task_id}

    video-get-task

    Returns the current status of a video generation task. When completed, includes the output_url.

  • POST/v1/video/generate

    video-generate

    Starts a text-to-video generation job using LTX-2 19B (distilled). Returns a task_id immediately. Poll GET /v1/video/tasks/{task_id} every 3-5 seconds until it reports completed, then read the output_url for the finished MP4.

  • GET/v1/video/files/{filename}

    video-get-file

    Downloads a generated MP4 by filename (the basename of output_url).

Frequently asked questions about MasterA AI GPU Inference API

How do I authenticate with this API?

All requests go through the JoJ API gateway using your JoJ API key (X-JoJAPI-Key header). No separate account with MasterA AI is needed — subscribe to a plan and the gateway handles authentication to our backend automatically.

Which models and GPUs power this API?

Requests run on our own NVIDIA RTX GPUs (5090, 5080, 4070 Ti Super) in the EU — no third-party cloud. Available models: Flux.1-Schnell and SDXL-Turbo for text-to-image, and LTX-Video / LTX-2 for text-to-video. Each request lands on whichever GPU has capacity, so there are no queues.