pplx-7b-online

Perplexity's PPLX Online LLMs are advanced language models that provide accurate, up-to-date responses by leveraging real-time internet data. These models, launched on November 29, 2023, are accessible via the pplx-api and Perplexity Labs.

Key features:
- Real-time access to current internet data, enabling accurate answers to time-sensitive queries
- Fine-tuned for effective use of web snippets, ensuring responses are current and factual
- Fast inference speeds using NVIDIA H100 GPUs, suitable for high-demand applications

How it works:
Users interact with the models through the pplx-api, which supports a RESTful interface for easy integration into various applications. Perplexity Labs provides a playground for experimenting with the models and refining queries. Users generate an API key through their Perplexity account settings to authenticate requests.

Integrations:
The API integrates with popular open-source models like Mistral and Llama, providing flexibility for developers. It is designed to be OpenAI client-compatible, facilitating easy integration with existing applications that use OpenAI's models.

Use of AI:
The PPLX models leverage the robust capabilities of `mistral-7b` and `llama2-70b` base models while enhancing them with real-time data access. Perplexity's proprietary search, indexing, and crawling infrastructure ensures the models have access to the most relevant and up-to-date information.

AI foundation model:
The PPLX models are built on top of `mistral-7b` and `llama2-70b` open-source base models.

How to access:
The PPLX Online LLMs are available as a REST API, making them accessible for web and mobile applications. Perplexity Labs provides a web-based interface for interacting with the models. The models are ideal for developers looking to integrate advanced LLM capabilities into their applications, businesses needing accurate real-time information, and researchers requiring precise and current data.