Together at TC Disrupt: 'token maxing without cost maxing' via smaller models
togethercompute · x · 2026-11-03
Speaking at TechCrunch Disrupt, Together AI's Vipul Ved Prakash argued the real goal is 'token maxing without cost maxing' — smart teams are swapping closed frontier models for smaller, faster open models to maximize output without ballooning costs, consistent with Together's open-source inference strategy.
More from Companies & People
- Dev: Gemini's G-Suite tools broke while Claude Code and Codex got more reliable — 3scorciav · 2026-11-03
- VibeCamp awarded a $300k grant from Lightcone Commons — amplifiedamp · 2026-11-03
- You.com CEO says the Hugging Face incident was 'a failure on OpenAI's side' — RichardSocher · 2026-11-03
- Kyunghyun Cho launches Ortet, a full-stack frontier AI lab arguing 'AI can't engineer medicine' — kchonyc · 2026-11-03
- Palantir's Karp: The US-China AI Race Is Defying Both Countries' Stereotypes — teortaxesTex · 2026-11-03
- Benj Edwards, Ars Technica's Founding AI Reporter, Is Job Hunting — genmon · 2026-11-03