Meter: Pay per thought AI

I'm Syed, one of the makers of Meter.

Working with AI apps over the last few years I faced four major problems.

1. I was paying hundreds of dollars a month for subscriptions to Claude, GPT, Grok, and Gemini and still hitting rate limits on all of them.
2. I would find myself copy pasting answers from GPT into Claude for a second opinion all the time.
3. I would jump straight into Claude Code / Cursor / Codex with a half formed idea, and end up with technical debt.
4. I would hesitate discussing new ideas with closed source models, but still had to because they are so much more intelligent than open source ones.

We built Meter to solve these problems, initially for ourselves, then realized this could be helpful for others.

Meter is the first pay-per-thought AI that debates itself.

1. Think first, pay later. Your balance accumulates as you use it, like a utility bill, and auto-settles when it hits the threshold. Meter auto routes so you never get rate limited. You see exactly what every message costs in real time, and can set caps and limits per message, per day or per month.

2. Watch agents debate each other in real-time. Get GPT, Claude, Grok, Gemini and others to debate your ideas and pushback using first principles. Research has shown that getting AI models to debate each other reduces hallucinations and increases the quality of responses.

3. Lock and track decisions over time. Turn your decisions into structured documents — readme.md, blueprint.md, design.md, decisions.md, claude.md. Export, share, or pipe them directly to your coding agents through our MCP server.

4. Chat privately on an anonymized basis. No email required. No password. Just a passkey on your device. Your identity is a random ID — we never see your name, and have no way of knowing who you are. Your thoughts remain truly yours.

The more we used Meter to build Meter, the more we found ourselves coming back to it. Debate mode, decisions log and agent specs make Meter significantly smarter than just GPT or Claude on their own. Think of Meter as ‘Git for thought’, the strategic layer that sits before you start coding.

Try it out and let us know which part you find most compelling, and what you’d like to see next.

Now in public preview at https://meter.chat.

Forum Threads

p/meter-2

•

5mo ago

Does getting AI to debate each other reduce hallucinations?

The research is surprisingly strong in favor of this.

Du et al. (ICML 2024) showed that when you get multiple LLMs to debate each other, not just share answers, but actively challenge each other's reasoning, then factual accuracy goes up and hallucinations go down. The wildest finding was that in some cases, every model started with the wrong answer but converged on the correct one through debate. The process itself generated correctness that no individual model had.

This is why xAI shipped a 4-agent debate inside Grok 4.20 three days ago. One of the leading AI labs looked at every way to improve output quality and landed on structured debate.

View all