Does Letta Message currently support image inputs?

[Azhis]

With the rise of multimodal models, does Letta support image inputs during chats? I didn’t see any mention of it in the documentation.

[lemorage]

As of now, I don't recall any specific plans for multimodal support in Letta, and it seems to me that, the current memory management is mostly focused on textual data. However, they have some examples that combine multimodal agents with Letta's memory management, like this one: letta-voice on GitHub. Just my two cents. @cpacker might have more insights on this.

[cpacker]

Hey @Azhis ! Thanks for your question - we are actively working on multi-modal and should have some updates coming soon on it. If you have a specific usecase in mind (even if it's as simple as "I sometimes want to share images with an agent and have them understand it"), please let us know since it helps drive our feature roadmap / helps inform our feature design.

[jd-jedi]

Is this available?