AI Panel Selects Hyperliquid, Solana and Zcash as Year-End Picks

A 10-model crypto exercise selected Hyperliquid’s HYPE, Solana’s SOL and Zcash’s ZEC among its leading year-end picks, offering forecasts rather than verified evidence of future performance.

Each model was asked to choose two cryptocurrencies: one from the current top 10 and another from the broader liquid, investable market. The exercise emphasized recent performance, valuation, adoption, momentum, catalysts and risks.

HYPE received six selections, the highest total. The exercise’s detailed responses cited trading activity, automated HYPE purchases funded by fees and the possibility of regulated U.S. access for perpetual products as forecast catalysts.

Trading fees support community mechanisms, including Hyperliquid’s Assistance Fund, which uses automated purchases of HYPE. Approximately 99% of protocol fees were allocated to the fund, and about 44 million HYPE had been acquired by April 30, 2026.

“Hyperliquid (HYPE) is the premier liquid, investable crypto asset for aggressive 2026 growth,” Qwen 3.7 Plus said in the exercise.

SOL received five selections. The models cited Solana’s planned Alpenglow consensus upgrade and possible institutional demand. Alpenglow targets approximately 150-millisecond finality, compared with about 12.8 seconds under the existing system, and was scheduled for expected mainnet activation in the third quarter of 2026.

ZEC received three selections. Grok described it as “the top-10 crypto most likely to deliver the biggest percentage gain by December 31.”

U.S. exchange-traded product developments also featured in the exercise. The Securities and Exchange Commission (SEC) declared the Bitwise NEAR ETF registration effective at 5 p.m. ET (21:00 UTC) on Sept. 24. The SEC declared a Grayscale-sponsored Zcash ETF registration effective Aug. 24, and the product was approved for listing on NYSE Arca.

A preliminary Form S-1 for a Winklevoss Zcash ETF was filed Oct. 6. The registration statement was incomplete, and shares could not be sold until it became effective.

The models were instructed to make aggressive, decisive calls and provide short answers. Their selections therefore represent AI-generated forecasts, not evidence that any asset will rise by Dec. 31.