Can you provide evidence the “more efficient” models are actually more efficient for vibe coding? Results would be the best measure.

Did I claim that? If so, then maybe I worded something poorly, because that’s wrong.

My hope is that as models, tooling, and practices evolve, small models will be (future tense) effective enough to use productively so we won’t need expensive commercial models.

To clarify some things:

• I’m mostly not talking about vibe coding. Vibe coding might be okay for quickly exploring or (in)validating some concept/idea, but they tend to make things brittle and pile up a lot of tech debt if you let them.
• I don’t think “more efficient” (in terms of energy and pricing) models are more efficient for work. I haven’t measured it, but the smaller/“dumber” models tend to require more cycles before they reach their goals, as they have to debug their code more along the way. However, with the right workflow (using subagents, etc.), you can often still reach the goals with smaller models.

There’s a difference between efficiency and effectiveness. The hardware is becoming more efficient, while models and tooling are becoming more effective. The tooling/techniques to use LLMs more effectively also tend to burn a LOT of tokens.

TL;DR:

• Hardware is getting more efficient.
• Models, tools, and techniques are getting more effective.