We asked two frontier AI models to predict Bitcoin's price by December 31st, they both refused to give a straight answer

Two AI Models, One Uncomfortable Consensus

@grok and @claudeai were each asked, in their top research modes, to give a single Bitcoin price target for December 31st. Neither would commit. SuperGrok's synthesis landed on a base case of $75,000 to $85,000, with a plausible range spanning $55,000 to $100,000-plus. Claude's Fable 5 placed its probability mass between $70,000 and $90,000, with a full range of $55,000 to $110,000. Two different systems, two different methodologies, and nearly identical answers.

Both models took the same position on why: the spread across real analyst forecasts, anywhere from a $25,000 to $50,000 bear floor to $150,000-plus bull targets, tells you more than any single number ever could. That view is consistent with what broader AI forecast experiments have found. Models tend to cluster around cautious ranges rather than bold calls, mapping uncertainty rather than resolving it.

When AI Gets the Facts Wrong

There was one notable slip worth flagging. Fable 5 cited Citi's current base case as $143,000 for $BTC. That figure is outdated. Citi had already lowered its Bitcoin target from $143,000 to $112,000 earlier in 2026. Then, on July 1st, the bank cut again. Citi reduced its 12-month price target for Bitcoin from $112,000 to $82,000. In a bear case scenario, the bank values Bitcoin at $53,000 over the next year.

Citi said it was forced to lower its forecasts due to three factors: lower investor appetite, ETF outflows, and a lack of progress on U.S. crypto legislation. U.S. spot Bitcoin ETFs recorded $4.5 billion in net outflows in June, their worst month since the products launched in January 2024. The AI model was working from a stale data point, which is a reminder that even sophisticated frontier models require source verification when applied to fast-moving markets.

The broader takeaway from the experiment is straightforward. Both AI models, drawing on wide bodies of analyst research, converged on a $70,000 to $90,000 zone as the weighted center of gravity for Bitcoin by year-end. That range happens to sit close to Citi's revised $82,000 base case. Nobody knows. But that is roughly where the probability mass lands once you strip out the noise at both extremes.

Sources
Citi cuts Bitcoin and Ether targets as ETF outflows deepen (Crypto.news)
Citi drastically slashes Bitcoin, Ether price targets (TheStreet)
14 AI Models Including Claude, ChatGPT and Grok Predict Bitcoin's Price Outlook (Bitcoin.com News)