Research results, launches and what changed in the product, with dates and numbers.
Research
OpenAI, Anthropic and Google lead among model labs, with Z.ai, Moonshot, DeepSeek and Alibaba from China. Rogue, from Brazil, has the top published score on two agent tests and sits above the public leaderboard on the third.
Research
Who competes on international leaderboards, who published new research this year and who is most cited and downloaded worldwide. Five labs, with the rule and the sources in the open.
Product
Describe the game and Rogue delivers a playable version in the browser. The games page has three games built by Rogue Constellation to play right now.
Research
The router reads the task before the agent starts and picks the reasoning effort. On Terminal-Bench 4.0 it matched high effort at 15% lower cost.
Product
In the race measured on October 31, ChatGPT in Thinking mode took 7 min 39 s for the same request. Rogue published its presentation guide with 16 uses.
Product
A website made on Rogue can open on your own domain, including on the free plan. A TXT record proves ownership and the verified site can appear on Google.
Research
Rogue's most advanced coding agent puts five agents on every task and returns the solution with the strongest evidence that it works.
Research
Rogue published the aggregates behind the DeepThink Preview charts as CSV, JSON and BibTeX, for anyone to check and cite.
Product
Speak or type a request and Rogue builds and revises the project in a folder on your Mac, with files and a preview beside the conversation.
Company
Rogue's site now exists in Portuguese at /pt-br, with the same pages as the English version. Visitors from Brazil see Portuguese first.
Research
Rogue's deepest reasoning mode builds on GLM 5.3 and gains 6.19 points over its own base on DeepSWE v1.1, ahead of GPT-6 Astra.