Poetiq Outperforms AI Giants on ARC-AGI-2 Using GPT Consortium

Poetiq has redefined the future of artificial intelligence by achieving state-of-the-art performance on the ARC-AGI-2 reasoning benchmark—not with its own proprietary model, but by strategically combining GPT and other third-party large language models. This breakthrough, accomplished by a mere six-person team, challenges the long-held assumption that only tech giants with massive internal AI infrastructure can lead in AGI development. Poetiq surpassed Google’s Gemini 3 Pro on ARC-AGI-2 while operating at half the cost and using only a fraction of the computational resources, signaling a seismic shift in how AI systems are designed and deployed.

A New Standard in AI Reasoning

ARC-AGI-2 is widely regarded as one of the most rigorous benchmarks for evaluating general reasoning capabilities in AI. Unlike traditional QA tests, it demands abstract thinking, novel problem-solving, and adaptation to unseen scenarios—skills closely aligned with human-like intelligence. Poetiq’s system doesn’t rely on a single model. Instead, it orchestrates a dynamic consortium of GPT variants, Claude, and open-source LLMs, each contributing their unique strengths. One model excels at mathematical deduction, another at symbolic logic, and a third at contextual inference. Poetiq’s proprietary orchestration layer intelligently routes queries to the most suitable model, then synthesizes outputs into a coherent, high-accuracy response. This hybrid architecture effectively neutralizes individual model weaknesses, creating a system more robust than any single LLM.

Small Team, Monumental Impact

Poetiq’s success demonstrates that dominance in AI no longer requires billion-dollar R&D budgets. The startup focused on algorithmic innovation, model alignment, and output optimization rather than model creation. By leveraging publicly available models, Poetiq reduced costs dramatically while increasing transparency and accessibility. This approach turns the AI race from a ‘model ownership’ contest into a ‘system orchestration’ contest. For emerging AI labs and startups worldwide, Poetiq offers a replicable blueprint: excellence is achievable through intelligent integration, not just scale. The implications extend beyond profit margins—it suggests a more democratic, decentralized future for AI innovation.

Poetiq’s achievement marks the dawn of a new AI paradigm: the era of the consortium. No longer is raw model size the ultimate metric; it’s the wisdom of combination. As the industry shifts from monolithic models to modular, collaborative systems, Poetiq stands not just as a startup—but as the vanguard of a smarter, more efficient AI future.