Home ยป NVIDIA Launches API Service for Running Models to Empower Developers in Building AI Applications for Free with Limitations

NVIDIA Launches API Service for Running Models to Empower Developers in Building AI Applications for Free with Limitations

NVIDIA has launched the NVIDIA API service, which brings together multiple popular artificial intelligence models, including NVIDIA’s own models, and opens them up as APIs for developers to use for free with limitations, focusing on developers who want to build prototypes as the main target.

The models provided for service fall into various groups:

LLM group models such as Gemma, Mistral, Llama2
StabilityAI image creation model
NVIDIA’s own embedding and reranking data search model
Image and text models such as Fuyu-8B, Google Deplot, Microsoft Kosmos 2, NVIDIA Neva 22B
Health-related models including specialized models like NVIDIA DeepVariant for finding differences in datasets, NVIDIA VISTA-3D for separating organs from CT images
Path optimization model NVIDIA cuOpt for logistics tasks
Weather forecasting model

The service is available for free with limited usage, providing 1,000 model run credits with the agreement not to use the data for further model training. However, additional credits are not available for purchase. Those who are interested in serious usage have the option to purchase NVIDIA NIM for continued usage.

Source – NVIDIA

TLDR: NVIDIA has introduced NVIDIA API, offering popular AI models as APIs for developers to use for free with limitations, catering to those focusing on prototype development.

More Reading

Post navigation

Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Meta Unleashes Llama 3 to Outsmart GPT-3.5, Claude 3 Sonnet, Gemini Pro 1.0 on the Horizon with 400B Model in Development

Solving Claude’s Reluctance: Unearthing Relevant Textual Predecessors amidst Anthropic Conundrums

Elon Musk Returns to Sue OpenAI and Sam Altman Once Again, Seeking Judgement on GPT-4 as AGI or Not