PPLX Models by Perplexity Labs
Details
Links
WebsiteAbout
The PPLX family of large language models by Perplexity AI is designed to tackle the challenges of traditional LLMs, such as dealing with outdated information and ensuring response accuracy. It features the core models pplx-7b-online and pplx-70b-online, which have 7 billion and 70 billion parameters, respectively 125. These models can access real-time data from the internet, enabling them to generate responses that are both current and factual 156. Building on open-source models like mistral-7b and llama2-70b, Perplexity's unique fine-tuning and search technology enhance these LLMs' efficacy 156. Evaluations indicate these models yield performance on par with or even surpassing prominent models like GPT-3.5 and Llama 2 in delivering precise and timely answers 157. Additional offerings include the pplx-7b-chat and pplx-70b-chat models, available via API and the Labs platform 156.
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Use when the workload needs 16k context and 70B parameters.
Use when the workload needs 16k context and 7B parameters.
Use when the workload needs 16k context and 70B parameters.
Use when the workload needs 16k context and 7B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Perplexity pplx-embed-v1 | Use when the workload needs 4B parameters. | 2024-05 | 4B parameters | Current |
| PPLX 70B Online | Use when the workload needs 16k context and 70B parameters. | 2023-11 | 16k context70B parameters | Current |
| PPLX 7B Online | Use when the workload needs 16k context and 7B parameters. | 2023-11 | 16k context7B parameters | Current |
| PPLX 70B Chat | Use when the workload needs 16k context and 70B parameters. | 2023-11 | 16k context70B parameters | Current |
| PPLX 7B Chat | Use when the workload needs 16k context and 7B parameters. | 2023-11 | 16k context7B parameters | Current |
| PPLX 8x7B Chat | Use when the workload needs 32k context. | 2023-11 | 32k context | Current |
| PPLX 8x7B Online | Use when the workload needs 32k context. | 2023-11 | 32k context | Current |
Release Timeline
2 release groupsSpecifications(7 models)
| Model | Released | Context | Parameters |
|---|---|---|---|
| Perplexity pplx-embed-v1 | 2024-05 | — | 4B |
| PPLX 70B Online | 2023-11 | 16k | 70B |
| PPLX 7B Online | 2023-11 | 16k | 7B |
| PPLX 70B Chat | 2023-11 | 16k | 70B |
| PPLX 7B Chat | 2023-11 | 16k | 7B |
| PPLX 8x7B Chat | 2023-11 | 32k | 8x7B |
| PPLX 8x7B Online | 2023-11 | 32k | 8x7B |
Frequently Asked Questions
- What is PPLX used for?
- The PPLX family of large language models by Perplexity AI is designed to tackle the challenges of traditional LLMs, such as dealing with outdated information and ensuring response accuracy.
- How does PPLX compare to Claude Fable?
- PPLX by Perplexity Labs is strongest where you need its listed use cases, while Claude Fable by Anthropic is the closest related family to check for vision and multimodal work. PPLX has 7 listed variants and reaches up to 32k context, while Claude Fable reaches up to 1m context, so compare the specs and pricing tables before choosing a production model.
- Which PPLX model should I use?
- If price is the main constraint, use the pricing table first because PPLX does not have complete provider pricing in the local data. For the most capable/latest local choice, evaluate PPLX 8x7B Chat with 32k context.
