All writing
·2 min·
ModelsWorkflowCost

Model loyalty is a trap

After hundreds of thousands of tokens this year — there's no one model to rule them all.

Hot take after burning through hundreds of thousands of tokens this year: model loyalty is a trap.

There's no one model to rule them all. No single agent. Switching is the skill now, not a sign you haven't figured it out yet.

The actual edge: knowing what's new and knowing what already works. Routing through OpenRouter, tracking real cost and latency through Helicone, picking the cheapest model that still passes evals on a given task. Switching like other people switch tabs.

If your stack assumes one provider, you're paying a tax — in dollars, in lock-in, and in being late to the next model that's better at your specific task.

Also published on, or originally from:

LinkedIn

LET'S TALK

Hiring? Or stuck mid-build?

Open to senior in-house roles with conversational AI / voice agent / AI-native DXP teams. Available for select Toptal engagements. I reply within 24 hours with something concrete — not a calendar link to a discovery deck.

→ Hiring conversations: tell me about the role and the stack

→ Project work: tell me the bottleneck, not the requirements doc

→ Either way, I'll come back with a take, not a pitch

LinkedIn DM

The fastest way to reach me. I'm already there every day — posts, comments, replies. Send a message and I'll come back within 24 hours.

/in/apsolut

GitHub

Pinned tools, side projects, the README that explains the vibe. CIA lands here when it's open-sourced.

/apsolut

Not ready to reach out?

Read what I've been writing about agents, model routing, and prompting. The fastest way to figure out if our wavelengths match.

Prefer email? Drop a line on LinkedIn first — I'll share the right address depending on whether the conversation is hiring, freelance, or something else.