Perplexity open-sources pplx-embed-v2-late: late-interaction embeddings for text, images, and pages
antoine_chaffin · x · 2026-10-09
Perplexity released pplx-embed-v2-late, two open late-interaction embedding models that retrieve text, images, and pages in a shared embedding space for cross-modal querying. The company claims frontier-level performance; both models are available on Hugging Face. Commentators framed it as a continued commitment to open-sourcing SOTA models.
More from Models
- Inception's Mercury Decide matches frontier accuracy on claim verification at lowest cost yet — StefanoErmon · 2026-10-09
- OpenAI bots hit 160K fetches for nonexistent URLs in a week, sparking RL-run speculation — gaganghotra_ · 2026-10-09
- Anthropic impresses on efficiency but users feel "babysat" by its guardrails — kimmonismus · 2026-10-09
- LightOnOCR-3 adds visual grounding and image description to document parsing — IgorCarron · 2026-10-09
- COLM paper: "forks in the road" in post-training data shrink reasoning model coverage — ParshinShojaee · 2026-10-09
- Google's open medical VLM MedGemma publishes in Nature Medicine — ymatias · 2026-10-09