The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestGoogle DeepMind2
Google DeepMindEmbeddings

EmbeddingGemma 2 Expands to Any Modality

Google DeepMind's compact embedding model now spans text, vision, audio, and video in a single vector space.

Sep 14, 2026
NotableGemma
EmbeddingGemma 2

Google DeepMind has released EmbeddingGemma 2, an embedding model built to turn text, images, audio, and video into vectors that applications can search, cluster, and compare. It arrives under the Gemma license and sits in the sub-1B parameter class, a size chosen to run close to where data lives rather than only in the cloud.

Embeddings are the quiet infrastructure behind retrieval-augmented generation, semantic search, recommendation, and deduplication. What sets this release apart is its reach across modalities: instead of stitching together separate encoders for each data type, EmbeddingGemma 2 aims to place different kinds of content into a shared representation, making cross-modal retrieval—finding an image from a text query, or a clip from a description—more straightforward.

Why it matters

  • A compact footprint makes on-device and private deployments more practical, reducing the need to ship raw data to a server.
  • One model across text, vision, audio, and video simplifies pipelines that today juggle multiple specialized encoders.
  • The permissive Gemma license lets teams build and ship without the friction of closed, API-only embedding services.

Google has not published parameter counts, embedding dimensions, or benchmark figures in the record we reviewed, so teams will want to validate quality on their own data before committing. Still, the combination of a small model and genuine multimodal coverage is a notable addition to the open embedding landscape, where most strong options remain text-first.

The model is available now on Hugging Face.

Sources

  • google/embeddinggemma-2

    Hugging Face

    Visit
OlderQwen-Image-2.1 Adds RGBA to Image GenerationQwen · Alibaba · Text → Image · 2 months agoNewerNari Labs Ships Qwen3-Based TTS and ASR ModelsUnknown · Text → Speech · 2 months ago

Get the model

Hugging Face

Specs

Size1.5 GB
PrecisionBF16
ArchitectureEmbeddingGemma2Model
LicenseGEMMA
Downloads610K
Likes3.4K

Can you run it?

Runs on just about anything.

  • BF16 (as published)

    Any modern laptop · 8 GB RAM

    2.5 GB
  • FP8

    Any modern laptop · 8 GB RAM

    1.7 GB
Your machine

Based on Hugging Face weights metadata. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

EmbeddingsAny-to-Any

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More from Google DeepMind

All Google DeepMind releases →
Google DeepMind/Any-to-Any

Google DeepMind's Gemma 4 Goes Multimodal and MoE

The new open-weights family adds a mixture-of-experts design, encoder-free multimodal inputs, and an optional thinking mode.

Jul 1, 2026
TabFM 1.0.0
Google DeepMind/Any-to-Any

Google DeepMind Releases TabFM for Tabular Data

A new foundation model brings zero-shot, in-context learning to classification and regression on structured tables.

Jun 29, 2026
DiffusionGemma
Google DeepMind/Text / LLMConsumer GPU

Google Releases Open-Source DiffusionGemma 26B Model

The new 26B parameter model from DeepMind uses a diffusion-based architecture, a technique more common in image generation, to produce text.

Jun 9, 2026

More in Embeddings

All Embeddings →
OpenBMB/Embeddings

NeoMME: a single-tower multilingual multimodal encoder

A new open encoder aims to make document retrieval faster by treating text and images natively in one model.

Sep 3, 2026
H company/Embeddings

H Company's NeoMME rethinks visual document retrieval

A single-tower multimodal encoder aims to make multilingual document search cheaper to fine-tune and run.

Aug 30, 2026
WeMM
Tencent/EmbeddingsRuns on a laptop

Tencent's WeMM-Embedding-9B Unifies Text, Image and Video

The WeChat team releases a 9-billion-parameter multimodal embedding model that maps three modalities into one shared vector space.

Aug 25, 2026