NEAR Protocol has launched a new feature enabling users to stake NEAR tokens in exchange for confidential AI inference access to more than 40 AI models from providers including Anthropic and OpenAI. The announcement was made via NEAR Protocol's official X account.

Under this new system, users retain ownership of their staked NEAR tokens while utilizing the AI inference services. This staking-based access model differs from conventional API pricing structures offered directly by AI providers, where users typically pay per token or per request in fiat or stablecoin currencies. By using staked tokens as an access mechanism, NEAR enables users to effectively pay for AI services through token yield or staking rights without liquidating their holdings.

The feature is effective immediately and represents a notable step in NEAR Protocol's broader strategy to integrate artificial intelligence capabilities into its blockchain ecosystem. NEAR Protocol co-founder Illia Polosukhin brings deep AI research credentials to the project, having co-authored the seminal 2017 paper "Attention Is All You Need," which introduced the transformer architecture underlying modern large language models from companies like Anthropic and OpenAI.

NEAR Protocol is a layer-1 blockchain platform designed for building decentralized applications and services, with an emphasis on scalability and developer usability. The platform has been actively pursuing the convergence of blockchain and AI technologies, including initiatives around chain abstraction and user-owned data.

The staking feature covers AI models from leading companies in the artificial intelligence space. The "confidential inference" framing indicates that user queries and responses are designed to remain private, addressing a concern that has grown as enterprises and individuals increasingly weigh the data privacy implications of sending sensitive prompts to centralized AI providers. By allowing users to access these models through staked tokens rather than direct payment, NEAR Protocol aims to increase the utility of its native token while providing users with continued exposure to their staked assets.

This development comes amid a broader industry trend of integrating AI technology into blockchain platforms. Several crypto projects have been exploring ways to bridge the two technologies, from decentralized AI compute marketplaces to token-based access for machine learning services. Projects like Bittensor have built decentralized networks for machine learning model training, while others have focused on decentralized GPU compute provisioning.

As of the announcement, specific market data regarding NEAR token trading activity in immediate response to the news was not readily available. The broader cryptocurrency market has been displaying mixed signals across various digital assets.

NEAR Protocol's initiative reflects the competitive landscape among blockchain platforms seeking to differentiate themselves through AI integration. The platform's ability to offer access to models from major AI providers through a staking mechanism positions it within the growing intersection of decentralized infrastructure and artificial intelligence services. A key factor to watch going forward will be user adoption metrics, including how many NEAR holders actively use the staking-for-inference feature and whether the model attracts developers building applications on top of the access layer.