Voyage Multimodal 3.5
Awaiting activationRetrieve evidence from charts, tables, slides and visual video content.
What it does
Voyage Multimodal 3.5 is Voyage’s current multimodal embedding generation. It represents interleaved text, images and visual video content in one embedding space, so a text question can find a chart, document screenshot or slide without relying entirely on OCR.
The official model guide lists a 32,000-token window and 256, 512, 1,024 or 2,048 dimensions. It adds video support over Multimodal 3. This service implements the native multimodal endpoint through AIRuntime; media access and commercial pricing are pending. Audio is not a supported input.
Content types
text · image · video
Find evidence in charts, tables, slides, document screenshots and visual video content. Does not embed audio.
AIRuntime media adapter implemented; commercial media activation is pending. The media API awaits commercial activation.
Where it helps
Financial reporting
Find the relevant chart or table inside report page images.
Visual business knowledge
Retrieve slides, diagrams and scanned-page screenshots by a business question.
Video discovery
Compare visual clips with a text or image query; this does not index their soundtrack.
When to choose it
Choose it when the evidence lives in visual documents and figures. For recordings or music, choose an audio-capable model instead.
Media adapter implemented in AIRuntime; commercial activation, provider credentials and verified pricing are required before calls are enabled. Inline payloads are bounded to 1 MiB per request; remote media URLs and large-file processing are not exposed. Always use a compatible model, task and dimension for indexed content and queries.
Official sources
Paraphrased from the provider’s documentation and original model information. Selection advice is our editorial guidance.
Voyage Multimodal 3.5 · Official model documentation ↗Reviewed October 4, 2026