NVIDIA: Comparing Models by Price Sheet Misses Half the Math, Token Efficiency Should Top Evals
togethercompute · x · 2026-11-10
Together AI shares a clip from Open by Design NYC: NVIDIA's Jacob Loewenstein argues that comparing models by their price sheet misses half the math, and that token efficiency belongs at the top of evaluation criteria whether you adopt a model or train your own. He appeared alongside Reflection's Zach Wentz and Together's Hassan El Mghari.
The post also promotes Together's upcoming events.
More from Models
- Dev finds "responsibility laundering" trick to bypass Opus refusals: delegate the dirty work to another model — corbtt · 2026-11-10
- One-shot prompt builds a full game in 8 hours for ~$550 of Claude API tokens — felixrieseberg · 2026-11-10
- Claude's hidden reasoning_effort tag exposed — and it secretly reads hidden messages — josh_bickett · 2026-11-10
- Opus Keeps Butchering a Joke-Rewriting Task, Ignoring Explicit Instructions to Preserve Punchlines — ivan_bezdomny · 2026-11-10
- Claude Fable 5.2 keeps butchering jokes: can't keep premise or punchline while rewriting — ivan_bezdomny · 2026-11-10
- Dev's early take on Fable 5.2: "best model," competitors will struggle to catch up — daniel_mac8 · 2026-11-10