# Perplexity — Multimodal embeddings beyond a single vector

- Company: Perplexity (perplexity.ai)
- Announced: 2026-10-07
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://www.perplexity.ai/hub/blog/multimodal-embeddings-beyond-a-single-vector
- Record: https://forck.live/items/16412-multimodal-embeddings-beyond-a-single-vector
- Subject: Perplexity
- Open weights: yes
- Models affected: pplx-embed-v2-late, pplx-embed-v2-context

Perplexity introduced pplx-embed-v2-late, a family of late-interaction embedding models that combine multi-vector representations, support for both text and image modalities, and a shared embedding space across model sizes (0.6B and 9B). The models achieve state-of-the-art results on vision and text retrieval benchmarks, with the 0.6B variant matching models five times larger on the ViDoRe benchmark, and are publicly available on Hugging Face.

## Evidence

Verbatim from https://www.perplexity.ai/hub/blog/multimodal-embeddings-beyond-a-single-vector:

> These new models achieve state-of-the-art results on a range of vision and text retrieval benchmarks, as well as agentic question-answering tasks. Notably, on ViDoRe (V3) , a leading visual document retrieval benchmark, our 0.6B model matches models with five times as many active parameters.

---

Record: https://forck.live/items/16412-multimodal-embeddings-beyond-a-single-vector
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md