Sakana AI

Sakana AI

Japanese AI company

Building Frontier AI in Japan

Founded: 2023
Based: Tokyo
Site: sakana.ai

@SakanaAILabs on X Wikipedia ↗

Last round

Raised: 32 billion yenSeries B
Announced: 16 November 2025

9 Jul

Learning Multi-Agent Coordination via Sheaf-ADMMResearch paper4 Jul

Bridging Spherical Black-Box OptimizersResearch paper3 Jul

Sakana Fugu: One Model to Command Them All21 Jun

Introducing Sakana AI’s Recursive Self-Improvement (RSI) Lab4 Jun

DiffusionBlocks: Training Neural Networks One Block at a TimeResearch paper27 May

Sparser, Faster, Lighter Transformer Language ModelsResearch paper8 May

KAME: Tandem Architecture for Enhancing Knowledge in Real-Time Speech-to-Speech Conversational AIResearch paper28 Apr

Learning to Orchestrate Agents in Natural Language with the ConductorResearch paper26 Apr

Trinity: An Evolved LLM CoordinatorResearch paper25 Apr

Sakana Fugu: A Multi-Agent Orchestration System as a Foundation ModelAvailability change23 Apr

String Seed of Thought: Prompting LLMs for Distribution-Faithful and Diverse GenerationResearch paper20 Apr

Digital Ecosystems: Interactive Multi-Agent Neural Cellular AutomataResearch paper18 Apr

The AI Scientist: Towards Fully Automated AI Research, Now Published in NatureResearch paper25 Mar

Instant LLM Updates with Doc-to-LoRA and Text-to-LoRAResearch paper26 Feb

Announcing a Strategic Partnership with DatadogPartnership26 Feb

Announcing a Strategic Investment from CitiCompany news24 Feb

Salesforce Ventures invests in Sakana AICompany news19 Feb

Office Relocation Announcement15 Feb

Sakana AI Announces Strategic Partnership with GooglePartnership23 Jan

An Unofficial Guide to Prepare for a Research Position Application19 Jan

RePo: Language Models with Context Re-PositioningResearch paper18 Jan

Extending the Context of Pretrained LLMs by Dropping Their Positional EmbeddingsResearch paper11 Jan

Digital Red Queen: Adversarial Program Evolution in Core War with LLMsResearch paper7 Jan

Sakana AI Agent Wins AtCoder Heuristic Contest (First AI to Place 1st)Research paper4 Jan

Announcing Our Series BCompany news16 Nov

Petri Dish Neural Cellular AutomataResearch paper4 Nov

Applied Team Member Introduction29 Oct

ShinkaEvolve in Action: How a Human-AI Partnership Conquered a Coding Challenge15 Oct

Sakana AI and Daiwa Securities Group to Develop AI for Advanced Asset ConsultingPartnership3 Oct

ShinkaEvolve: Evolving New Algorithms with LLMs, Orders of Magnitude More EfficientlyDeveloper tool24 Sep

Competition and Attraction Improve Model FusionResearch paper24 Aug

Bringing cutting-edge AI technology to business: Interview with Sakana AI Applied Team members28 Jul

Inference-Time Scaling and Collective Intelligence for Frontier AIResearch paper30 Jun

Reinforcement Learning Teachers of Test Time ScalingResearch paper22 Jun

The last 8 weeks

Lead story

Recursive Self-Improvement through Collective Intelligence

The last 8 weeks

Lead story

Recursive Self-Improvement through Collective Intelligence

Sakana AI

The last 8 weeks

Sakana AI

Last round

Sakana AI

Last round