Research3 min read

Perplexity launches Frontier Research to build an automated research scientist

Perplexity has started an open research effort aimed at building an AI system that can conduct research on its own, backed by more than 10,000 GPUs. It commits to releasing papers, data, models and code.

AE
Aidose Editorial
Nov 17, 2026

Perplexity announced Perplexity Frontier Research on November 16, a new initiative to advance algorithms, models and evaluations in the open. "We aim to build an automatic research scientist, an AI system that can conduct research on its own," the company wrote. The work will be published at research.perplexity.ai.

Perplexity says the effort is meant to produce interpretable knowledge that transfers across fields, rather than black-box predictions in narrow settings. It is organized around six areas: multi-agent orchestration, pre-training, continual learning, reinforcement learning, systems and evaluations. Example questions in the announcement include whether more efficient architectures can be proposed, whether synthetic data can support broader generalization, when multi-agent collaboration works best, and how to measure research ability and extrapolation rather than recall.

The company says the research will be supported by "considerable computational resources, in excess of 10,000 GPUs," and commits to releasing "papers, blog posts, data, models, and code." The announcement does not name the researchers leading the effort, give a budget or list initial projects. Perplexity is hiring full-time researchers and running a research fellowship program.

The move takes Perplexity, best known for its AI search product, further into foundational research. If it follows through on publishing models, data and code, its evaluation work on measuring research ability could be useful to teams building agents for scientific and technical tasks.

AE
Aidose Editorial
Nov 17, 2026 · 3 min read
Back to News