Continuum Provider Error: Not a multimodal model

[AnthonyRonning]

Investigation into logs reveals a 400 error when sending requests to llama-3.3-70b:
Provider continuum returned status 400 Bad Request: {"error":{"message":"/mnt/modelweights-textgen-0 is not a multimodal model None","type":"BadRequestError","param":null,"code":400}}

Context:

• Occurred in opensecret::web::openai.
• Model: llama-3.3-70b.
• Timestamp: 2025-11-23T03:46:36.504336Z.

Possible causes:

• Sending multimodal content (images, etc.) to a text-only model.
• src/web/responses/handlers.rs might not be filtering content types correctly for the target model.

We should ensure that multimodal content is only sent to models that support it.