Our general-purpose reasoning models solved all 12 problems at the 2025 International Collegiate Programming Contest (ICPC) World Finals, the world’s top university programming competition which was enough for a 1st-place human ranking.
RESEARCH
-
OpenAI o1 Models Achieve Top Marks in Math Coding Competitions
By
–
This caps a run of steady progress across math and coding competitions. Just over a year ago we introduced OpenAI o1-preview and OpenAI o1-mini. Since then our general-purpose reasoning models have made steady progress. Today they’re earning top marks in some of the world’s
-

Self-Correcting AI Agents Enable Exponential Task Horizon Gains
By
–
I think the significance of this is under-appreciated: the assumption has often been that AI agents are brittle as one failure in a chain breaks a task But this paper shows smart models are self-correcting & that small gains in accuracy lead to exponential gains in task horizons
-
Gemini 2.5 Deep Think Achieves Gold-Medal Performance at ICPC World Finals
By
–
Very excited to see our Gemini models getting better and better at coding! An advanced version of Gemini 2.5 Deep Think at the 2025 International Collegiate Programming Contest (ICPC) World Finals achieved gold-medal level performance!
-
Optimism Over Pessimism: Embracing Technology’s Bright Side
By
–
Better to live life erring on the side of being optimistic and wrong than pessimistic and right! Be realistic, but, as Monty Python would say, always look on the bright side of life!
-

OpenAI Wins Math Competition Solving All 12 Problems
By
–
Otra competición de matemáticas donde OpenAI participa y logra oro superando en este caso todos los problemas (12 de 12)! No han usado un modelo entrenado específicamente para esto sino un combo de modelos generales. Entre ellos el misterioso modelo avanzado de matemáticas
-
100+ Free Step-by-Step AI Agents and RAG Tutorials
By
–
100+ free step-by-step tutorials with code covering: AI Agents RAG Systems Voice AI Agents MCP AI Agents Multi-agent Teams Autonomous Game Playing Agents P.S: Don't forget to subscribe for FREE to access future tutorials.
-

Alibaba releases 30B agentic LLM outperforming Claude and DeepSeek
By
–
China's Alibaba just dropped an opensource 30B agentic LLM that outperforms Claude 4 Sonnet, DeepSeek v3.1, Kimi k2 on a range of agentic search benchmarks. Only ~3B parameters are activated per token. 100% Opensource.
-
Advancing Scheming Research in AI Preparedness Framework
By
–
This is significant progress, but we have more work to do. We’re advancing scheming research categories in our Preparedness Framework, renewing our collaboration with Apollo, and expanding our research team and scope. And because solving scheming will go beyond any single lab,
-

Frontier Models Show Situational Awareness Affects Scheming Behavior
By
–
Frontier models can recognize when they are being tested, and their tendency to scheme is influenced by this situational awareness. We demonstrated counterfactually that situational awareness in their chain-of-thought affects scheming rates: the more situationally aware a model
