Introduction

Moonez.ai provides native and Gemini-compatible APIs with a shared key and balance:

• Images — five Nano Banana models for text-to-image generation and image editing.
• Text — nine Gemini models billed per token, including Gemini 4 Argon and Gemini 3.8 Flash.
• Video — Veo 3.1, Veo 3.1 Fast, and Veo 3.1 Lite.
• Music — Lyria tracks and native RealTime audio streaming.
• Multimodal — Gemini Omni, with provider-dependent input and output capabilities.

Start with the complete catalog of 23 models. The latest additions are gemini-4-argon, gemini-3.8-flash, and gemini-3.1-flash-lite-image (Nano Banana 2 Lite, 1K only).

Base URLs

API	Base URL
Gemini-compatible API	https://api.moonez.ai/api/gemini/v1beta
Native jobs API	https://api.moonez.ai/api/v1

For Android apps with a custom Gemini provider, enter the complete URL https://api.moonez.ai/api/gemini/v1beta. Use your Moonez.ai key.

In the Python SDK, configure the same URL together with base_url_resource_scope=types.ResourceScope.COLLECTION to prevent the SDK from adding /v1beta again. See the SDK setup.

How to get results

The native API is asynchronous:

1. Send POST /api/v1/generations with the model and parameters.
2. Receive 202 Accepted, a job_id, and a status_url.
3. Poll GET /api/v1/jobs/{job_id} until the status is done or failed.
4. If the result contains an image, download it from result.image_url.

You can configure a webhook instead of polling.

The Gemini-compatible API returns the result directly in the generateContent response or through a streamGenerateContent stream. No job polling is needed for these methods. Veo uses asynchronous jobs instead. See the API overview.

Image models

For the native API, use POST /api/v1/generations with the model field:

Model	Native API resolution	Generation	Editing
gemini-2.5-flash-image	1K	Text-to-image	Image-to-image
gemini-3.1-flash-lite-image	1K only	Text-to-image	Image-to-image
gemini-3.1-flash-image	1K · 2K · 4K	Text-to-image	Image-to-image
gemini-3-pro-image	1K · 2K · 4K	Text-to-image	Image-to-image
gemini-nano-banana-2.1	1K · 2K · 4K	Text-to-image	Image-to-image

The same IDs are available through the Gemini-compatible API.

Text and video

See Text models for the model list and per-token billing details.

For veo-3.1-generate-001, veo-3.1-fast-generate-001, and veo-3.1-lite-generate-001, use POST /api/gemini/v1beta/interactions. Request formats and billing are covered in Video generation.

The multimodal gemini-omni-1.1-flash model is called through POST /api/gemini/v1beta/models/gemini-omni-1.1-flash:generateContent. See Models for details.

Next steps

• Quickstart — your first generation with cURL.
• Authentication — keys, permissions, and idempotency.
• Python SDK — Nano Banana 2.1 generation through the Gemini-compatible API.
• API Reference: Nano Banana 2 Lite — 1K generation and model-specific access.
• Model catalog — every active model and its exact API ID.