The open-source tool will assess AI models' capabilities to generate and execute transactions with Solana's SDKs and DeFi protocols, alongside new funding for research.
The Solana Foundation has launched Solana Bench, an open-source benchmark designed to evaluate how AI models such as Claude, GPT-5, and Gemini interact with Solana SDKs and DeFi protocols. The initiative also includes new funding for ongoing research in this domain.