The model behind your content should be your call.

Not every piece you make needs your most powerful model. A quick social caption and a flagship article are different jobs, and treating them the same either overpays on the easy work or underpowers the work that matters. So in Studio's Content Repurposer, the model is becoming a dial you control. And I'm going to show you exactly what each setting costs.

Most AI tools quietly pick a model for you and never mention it. That single choice is the biggest lever on quality, speed, and price, so hiding it leaves you with a one-size-fits-all tool. I'd rather hand you the lever, and the price tag that comes with it.

How the tool works, in one paragraph

You give Studio a real source. A sermon, a podcast, a chapter you wrote. It reads the whole thing, indexes it so it can pull the right passages, and then writes the piece you asked for using only what is actually there. The model is the engine doing that reading and writing. A fast, cheap model gets through a lot of work. A flagship model reasons harder and stays closer to your source. Same tool, different engine.

Stronger does not always mean more faithful

This one surprised me. I went looking for the best model and learned the question was wrong. The models that top the intelligence leaderboards are not always the most faithful to a source. The trait that actually matters is unglamorous: a model that does not invent claims, and that admits when your source does not cover something instead of inventing a tidy answer.

When researchers put the current models through real retrieval tests, Claude's came out ahead on exactly that. The lowest invention rate, and the highest rate of correctly saying the answer is not in the source rather than filling the gap. For a tool whose whole promise is faithfulness, I care about that far more than raw IQ. So the strongest setting on the dial runs Claude.

The lineup, and what each costs

Studio runs on credits, which are just metered usage. A light, fast model spends a fraction of a credit per piece, so a plan stretches to hundreds of them. A flagship model reads and reasons more, so it costs more credits per piece. That is the whole trade, and now you can see it before you spend, not after.

The dial will default to a fast, capable model, because most repurposing does not need a flagship and I'd rather you get real mileage out of a plan. When a piece truly matters, you turn it up, and the price moves with it in plain sight.

Showing you the engine and the meter

Straight with you: the model picker is live today with the everyday models, and the flagship tiers and the dial itself are rolling out now. I'm telling you what is coming because that is how I build, and because you should be able to see the engine and the meter, not just the output.

Power, speed, and price should be your call on each piece, not a decision I make for you. That is the bet, and it is the same one behind everything in Studio.