Multimodal Overview
NagaAI supports several input and output modalities. In addition to text, you can send images, documents, and audio files to our Chat Completions API. Certain models can also generate images.
API format
The request format matches the OpenAI Chat Completions API. Additionally, for file input, you may provide an http/https URL; we will download it, validate it, and forward it to the upstream provider. "File" input refers to PDF documents.
Model capabilities
You can discover input/output modalities for each model on the NagaAI Models page.
Pricing
Cost estimation is performed by the upstream provider in tokens. In the future, when models lack native support for these modalities, we may use external tools (e.g., OCR for PDFs) to enable them.