Does Gradient AI support vision / multimodal ?

Question

I’m looking to do image analysis (extract title, description and keywords) for JPG images.

With Anthropic Claude e.g., I know I can do this directly through their API. However, I’m interested in GradientAI, to add a knowledgebase e.g.

So my question is, will I be able to use vision/multi-modal in GradientAI and with which of the currently supported LLMs?

Thanks!

yolanda smith · Answer

Oh, I’ve been wondering the same! From what I’ve seen, Gradient AI mainly focuses on text-based tasks but i am not 100% sure about vision or multimodal support. Has anyone tried uploading images or combining text + image prompts with it? Curious if it actually handles that or not.

alexdo · Answer

Hello, To actually answer this rather than just wonder alongside you: yes, Gradient AI Platform has vision capable models you can select today, not just image generation. DigitalOcean’s model list documents several with native image input, Moonshot AI’s Kimi K3, K2.6, and K2.5, NVIDIA’s Nemotron Nano 12B v2 VL (built specifically as a multimodal variant), Z.ai ’s GLM-5.3 Flash which takes text, image, and video input, and Alibaba’s Qwen3.8-Max. Those are separate from the gpt-image-1 and gpt-image-1.5 models, which are text to image generation only and won’t help for your use case of extracting information from an existing JPG. Since an agent is really a configuration wrapped around a chosen underlying model, picking one of the vision capable models above as your agent’s model should let it accept image input the same way it would through direct serverless inference, worth confirming that specific combination (image input plus your knowledge base attached) behaves as expected in the playground before you build against it, since the docs describe the model capability and the knowledge base feature separately rather than as a single documented combination. Regards

Become a contributor for community

DigitalOcean Documentation

Resources for startups and AI-native businesses

The developer cloud

Start building today

Report this

Does Gradient AI support vision / multimodal ?

Related questions

Using NVIDIA GPUs for Scientific Research

Report this

Does Gradient AI support vision / multimodal ?

Related questions

Using NVIDIA GPUs for Scientific Research