For AI startups, selecting a model no longer represents a simple one-time architectural decision. Open models are improving rapidly, and frontier APIs continue to advance. Companies can now customize models for specific workloads, with some organizations building products that utilize multiple models rather than committing to just one. This evolution provides founders with more avenues for building, spending, and owning technology, allowing for increased flexibility as the industry evolves.
The session "The Real Tokenmaxxing: How the Best AI Companies Navigate a Multi-Model World" will convene Mo Jomaa, partner at CapitalG; Vipul Ved Prakash, co-founder and CEO of Together AI; and Zuzanna Stamirowska, CEO and co-founder of Pathway, on the Builder’s Stage. They will discuss why companies are adopting multi-model strategies, how they balance cost and performance, and when open models outperform proprietary alternatives.
Manos Koukoumidis, CEO and co-founder of Oumi, will lead a session on the Real World AI Stage titled "Which AI Should Your Company Actually Deploy: Rent, Customize, or Build." He will help attendees evaluate frontier APIs, customized open weights, and owning AI outright through practical frameworks and decision principles.
The choices founders make regarding model architecture can shape both the product and the business. Building more of the AI stack offers greater control and differentiation, but it can also require significantly more time and resources than relying on existing models. Attendees are encouraged to add the Oumi session to their Disrupt agenda to determine the appropriate level of AI stack ownership.
Nader Khalil, Director of Developer Tech at Nvidia, and Sydney Sykes, Global Head of VC Partnerships at Nvidia, will address the Builders Stage in "Building AI Startups Worth Betting On." They will analyze current founder choices regarding frontier APIs versus open-weight models and how these decisions impact product strategy and long-term differentiation.
Model architecture is only one part of the equation, as AI performance ultimately depends on the underlying hardware. Anna Goldie and Azalia Mirhoseini of Ricursive Intelligence will explore this intersection on the Disrupt Stage in "When AI Starts Designing Its Own Hardware." Their discussion will cover how AI is optimizing chips and hardware, the closer connection between architecture and hardware, and the implications for an increasingly open ecosystem.
For founders, the link between AI and hardware can determine how quickly new capabilities reach the market. Faster chip development could also alter the infrastructure available to startups building the next generation of AI products.
These conversations are part of over 200 sessions across six industry stages, roundtables, and breakouts at Disrupt 2026, taking place October 13-15 at the Moscone Center in San Francisco. With more than 10,000 founders, investors, and tech leaders expected, along with 250+ speakers and 300+ exhibiting startups, the event offers matchmaking and networking opportunities.
A startup may utilize a frontier API today, customize an open model tomorrow, or eventually move workloads among several models. Preserving flexibility remains as important as choosing the right model initially.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.