HarnessAgent now supports image inputs for all harness adapters

T1Covered by 1 source · first reported by Vercel (blog + changelog) at 27 Oct, 00:00 UTC

HarnessAgent The AI SDKlayer now accepts image inputs across every harness adapter. content { type: 'text' } { type: 'file' } Pass images alongside text in the standard AI SDK message format: aarray mixingandparts. Image input is processed by the harnesses' underlying vision models and supported…

dev

Sources

• T1HarnessAgent now supports image inputs for all harness adaptersVercel (blog + changelog)27 Oct, 00:00