StoryFirst seen
Perplexity says its models read a billion tokens in one day and led to 14 fixes in its agent security rules.
Everything it touches
2 names 2 voices 2 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
Neat example of how Perplexity used Decider internally In one day it read 1B tokens across 7,000 agent sessions and found flaws that led to 14 fixes to their agent security rules. Both models are Apache-licensed and free on OpenRouter until Nov 19Read original on X
Decision Models are the future of agentic security. They can process billions of tokens of telemetry, identify security issues, and help you patch fixes at a scale and cost structure not possible before. We’ve open sourced our models and made it free on Openrouter for a week.Read original on X
Perplexity open-sources its agent security models, free on OpenRouter
Perplexity open-sources agent security models, free on OpenRouter
Perplexity open-sources agent security models, free on OpenRouter for a week
Perplexity open-sources Decider agent security models, free on OpenRouter
Perplexity open-sources Decider agent security models, free on OpenRouter
Shared connections
The deal puts a model routing platform used by more than 10 million developers inside a major payments company, giving it control over AI costs and data location.
The model targets agentic workflows in knowledge work, software engineering, cybersecurity, and science.
It follows the Argon model by a few weeks and targets agentic work in coding, cybersecurity and knowledge work.
Muse Ember is pitched at a third of the cost or less of rival small models, aimed at repeated, token-heavy agent tasks.