Capping Claude Opus at 10 runs a day — and why that's a feature

Most platforms hide the ceiling and meter you out of it silently. We wrote ours on the pricing page: Quorum Max = 10 runs/day, hard, for everyone.

What a Max run actually does

Each Max ticker fires four elaborated frontier-model prompts in sequence:

• Deep research — the full bull thesis grounded in multi-year fundamentals + last quarter's filings.
• Short-seller — the strongest sell case with named risks, accounting flags, customer-concentration.
• Exit plan — under what conditions does this trade end? At what level, by what date, on what news.
• Consensus — frontier-model synthesis with at least two preserved dissents.

Output: a canonical Executive Briefing with TL;DR at the top, the bell-curve verdict, full per-seat reasoning. ~3-5 minutes async — you keep working while it runs.

The cap, plainly stated

• 10 runs per UTC day, across the whole instance.
• Cap applies to every caller including owners and the operator's own legacy token — no bypass.
• Complete ticker analyses from the last 15 days are reused; only genuinely new or expired work consumes compute.
• When you hit the cap you see a 429 with daily_used/daily_cap and the UTC reset time.

Why the cap is a feature

Three failure modes the cap prevents:

• Capacity protection. A single user (or a leaked token) hammering the frontier endpoint at 04:00 UTC degrades the service for everyone else watching the same queue. A hard cap is a hard ceiling on that damage — it protects shared infrastructure for every active user.
• "I ran it 47 times and didn't read any of them." Frontier verdicts cost real attention. 10 a day is enough for the calls you actually need to decide; less than 10 is enough for the calls you actually need to decide; 50 is just a way to avoid reading any one of them.
• Visible ceiling. Pricing pages with invisible caps are a tax on trust. We'd rather tell you the number.

See the full tier comparison →   Claim a seat →