Sakana Marlin adds Interactive Reading and PowerPoint export for deep research reports
SakanaAILabs · x · 2026-09-16
Sakana AI updated its Ultra Deep Research Assistant "Sakana Marlin" with two features targeting the bottleneck of understanding AI research output:
- Interactive Reading: chat with the agent about any section of a generated report, and clicking a citation makes the AI open the source and pinpoint the exact passage backing a claim — instead of users hunting through source pages.
- PowerPoint export: investigation results can be turned into editable slides for reporting.
Context: Marlin, built on the AB-MCTS technique, autonomously runs up to 8 hours of research from hypothesis to report, and is already used by enterprises, consultancies, financial institutions, and universities. Sakana says the features extend AI support to the "understand, verify, communicate" steps of decision-making.
More from coding & agent
- Running more than 2 sub-agents at once burns tokens for zero quality gain, dev argues — kevinnbass · 2026-09-16
- Together AI CPO cuts PRDs to 2 pages plus a prototype to fix "context window explosions" — aakashgupta · 2026-09-16
- LangChain shares how it built its Paid Media Agent — BraceSproul · 2026-09-16
- Notification tragedy of the commons: a personal agent should suppress ~95% of them — signulll · 2026-09-16
- skillmem: open-source procedural memory for MCP agents, with hard-won recursion lessons — mrPetrukovich · 2026-09-16
- Running More Than 2 Sub-Agents in Parallel Burns Tokens for Zero Quality Gain — pvncher · 2026-09-16