API Reference

This guide will help you get set up with the MegaNova APIs, organized by tier-specific features and limitations.

Getting Your API Key

API keys are automatically generated for each user upon account creation. You can view, renew, or revoke your API key in the MegaNova user dashboard:

• View/Renew API Key: Log in to your user dashboard to view or renew your API key. If you renew your key, the old one will be invalidated.

• Security Note: Keep your API key secure. If you suspect it has been compromised, renew it immediately.

API Types

MegaNova provides two main APIs:

Platform API

• For managing billing and account resources.

• BASE URL: https://api.meganova.ai

• BASE PATH: /api/v1

• API_URL: https://api.meganova.ai/api/v1

Inference API (OpenAI Compatible)

• For running inference (text, vision, image, embeddings) using serverless endpoints.

• BASE URL: https://api.meganova.ai/v1

• OpenAI compatible: Use OpenAI SDKs and tools.

Enterprise Endpoint (OpenAI Compatible)

• The same models and the same API key, served from dedicated enterprise deployments instead of the shared pool.

• BASE URL: https://enterprise.api.meganova.ai/v1

• See Enterprise Endpoint for which models are available and how usage is billed.

Tier-Specific API Access

🚀 Tier 1

• ✅ Free Models Only: Access to <100B parameter models

• ✅ Basic API Access: Standard inference endpoints

• ❌ Premium Models: No access to >100B models

• ❌ Dedicated Endpoints: No GPU hosting access

🎨 Tier 2

• ✅ Free Models: <100B parameters (including Manta Flash)

• ✅ Premium Models: Deepseek V3 (free), Paid models up to 5,000 RPD

• ✅ Basic Dedicated Endpoints: Access to 4090 and A6000 GPUs

• ✅ Standard Support: 72-hour response SLA

⚡ Tier 3

• ✅ Free Models: <100B parameters (including Manta Flash)

• ✅ Premium Models: Deepseek V3 (free), Paid models up to 10,000 RPD

• ✅ Enhanced Dedicated Endpoints: Access to L40S and A100 GPUs

• ✅ Priority Support: 24-hour response SLA

👑 Tier 4

• ✅ Free Models: >100B parameters (including Manta Flash)

• ✅ Premium Models: Deepseek V3 (free), Paid models up to 50,000 RPD

• ✅ All Dedicated Endpoints: Access to all GPU types including H200/H100

• ✅ Premium Support: 12-hour response SLA

Rate Limits by Tier

Tier	Text Generation	Image Generation	Dedicated Endpoints	Support SLA
Tier 1	Free models only	Basic	❌ Not available	Community
Tier 2	5,000 RPD	Standard	4090, A6000	72 hours
Tier 3	10,000 RPD	Enhanced	L40S, A100	24 hours
Tier 4	50,000 RPD	Maximum	All GPUs	12 hours

API Documentation

Platform API

• Authentication

• API Keys Management

• Billing Management

• Usage & Limits

• Errors & Response Conventions

Inference API

• List Models

• Generate Text

• Generate Images

• Generate Embeddings

• Vision Capabilities

See Also

• Tier Overview

• Pricing Information

• Glossary

• Legal Documentation