AI Lab Private Valuations 2026

Shared workspace on Qwidgets for tracking prediction markets across Kalshi, Polymarket, Manifold, and PredictIt. Live prices, candlestick charts, order books, and historical analytics for every event and market on this page.

Widgets

What this workspace tracks

On May 19, 2026 Polymarket launched the first prediction markets settled by Nasdaq Private Market (NPM) data. The launch contracts price binary thresholds on private company valuations — OpenAI, Anthropic, and others — at fixed snapshot dates. NPM supplies the institutional-grade transaction and valuation tape; the binary contract IS the price discovery mechanism for an underlying with no public reference price. Read the year-end and June 30 ladders for OpenAI and Anthropic side by side. Spotlight markets below add the "Anthropic $500B+ in 2026" single binary and the cross-vertical "OpenAI vs Meta higher valuation on December 31" pairwise contract.

Companion workspace for IPO-conditional payouts at /w/spacex-ipo-tracker.

Polymarket's December 31 strike ladders for OpenAI and Anthropic. Each strike is a one-touch binary on whether the company's NPM-resolved private valuation reaches that level by December 31, 2026.

Six weeks out. The June 30 ladder tightens distribution mass quickly as the snapshot date approaches; the slope of the strike chain is a clean read on how confidently the market has converged on a single valuation band.

The two highest-information single binaries in the launch suite — a one-touch Anthropic $500B threshold for full-year 2026, and the cross-vertical OpenAI-vs-Meta market-cap comparison at year-end (binary on which company is larger on December 31).

What to watch.

  • NPM is the settlement source, not the underlying. These binaries don't track a continuous public price — they resolve against NPM's transaction-based valuation marks at the snapshot date. The price you see is a discrete binary on a discrete resolution, not a continuous derivative on a continuously-quoted asset. That's structurally different from a stock-price prediction market.
  • Distribution slope on the strike ladders. A steep ladder (one strike near 0.5, neighbors near 0/1) means the market has a tight modal valuation. A flat ladder across multiple strikes means real tail dispersion. Both readings carry information regardless of where the central mass sits.
  • OpenAI vs Anthropic ladder spread. Where Anthropic's $500B strike trades vs OpenAI's $500B strike is the cleanest read on relative private-company positioning across the two leading AI labs. Watch divergence after major fundraising or product announcements.
  • OpenAI-vs-Meta pairwise binary. Inverts the public-vs-private framing — Meta's valuation is a continuous, publicly-quoted reference; OpenAI's is a discrete NPM mark. A 0.5 print on this binary means the market thinks the two companies' year-end valuations are roughly equal, modulo the day-to-day jitter of Meta's public quote. This is the first pairwise public-vs-private valuation contract that I'm aware of.
  • Companion IPO-conditional markets at /w/spacex-ipo-tracker. Those are conditional on an IPO event happening — pre-IPO valuation here, post-IPO closing market cap there. Reading them together gives a full pre/post-listing valuation arc.