---
schema: 1
type: video
title: 'Gemma, DeepMind''s Family of Open Models — Omar Sanseviero, Google DeepMind'
description: >-
  Google DeepMind’s Gemma family is expanding. Join us for a deep dive into the
  latest models of the Gemma ecosystem. From vibe fine-tuning to Sovereign AI,
  you'll learn about the latest model capabilities, how to build
  high-performance applications, and how to get started with open models.
  Speaker info: - https://x.com/osanseviero -
  https://www.linkedin.com/in/omarsanseviero/ - https://github.com/osanseviero
  Timestamps 0:00 Introduction to the Gemma model family 0:41 Evolution from
  Gemma 3 to Gemma 4 1:21 Overview of the new Gemma 4 capabilities 2:31 Live
  demonstrations of on-device applications 3:38 LM Arena scores and performance
  benchmarks 5:07 Apache 2 license transition 5:27 Technical deep dive: The E2B
  architecture and per-layer embeddings 6:57 Multimodal understanding and
  multilingual support 8:43 Ecosystem growth and community adoption 10:07
  Product integrations, including Android Studio 10:46 Statistics on model
  downloads and fine-tuning 11:27 Official Gemma variants: Shield Gemma and
  MedGemma 12:16 Community research and sovereign AI efforts 12:56 Real-world
  applications, from cancer therapy to offline tasks 14:05 Closing remarks and
  future outlook --- Watch along with the corrected transcript, summary,
  timestamps and resources on this talk's official AIE page:
  https://ai.engineer/talks/_gVFUEdhCyI-gemma-deepmind-s-family-of-open-models
  Join our upcoming AI Engineer conferences (see https://ai.engineer landing
  page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs
  and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference
  returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS
  2026 in Sydney ---
canonical: 'https://liquidgold.cc/watch/_gVFUEdhCyI'
markdown: 'https://liquidgold.cc/watch/_gVFUEdhCyI.md'
llms: 'https://liquidgold.cc/watch/_gVFUEdhCyI/llms.txt'
parent: 'https://liquidgold.cc/channel/aidotengineer.md'
channel: aidotengineer
updated: '2026-10-24T15:20:58.069Z'
language: null
children: []
videoId: _gVFUEdhCyI
durationSeconds: 926
publishedAt: '2026-04-20T15:15:06.000Z'
searchChunks:
  - ordinal: 0
    text: null
    startSeconds: 26
    endSeconds: 50
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=26'
    heldReason: rights
  - ordinal: 1
    text: null
    startSeconds: 82
    endSeconds: 93
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=82'
    heldReason: rights
  - ordinal: 2
    text: null
    startSeconds: 93
    endSeconds: 158
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=93'
    heldReason: rights
  - ordinal: 3
    text: null
    startSeconds: 93
    endSeconds: 158
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=93'
    heldReason: rights
  - ordinal: 4
    text: null
    startSeconds: 170
    endSeconds: 224
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=170'
    heldReason: rights
  - ordinal: 5
    text: null
    startSeconds: 310
    endSeconds: 320
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=310'
    heldReason: rights
  - ordinal: 6
    text: null
    startSeconds: 332
    endSeconds: 344
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=332'
    heldReason: rights
  - ordinal: 7
    text: null
    startSeconds: 344
    endSeconds: 423
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=344'
    heldReason: rights
  - ordinal: 8
    text: null
    startSeconds: 344
    endSeconds: 423
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=344'
    heldReason: rights
  - ordinal: 9
    text: null
    startSeconds: 423
    endSeconds: 433
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=423'
    heldReason: rights
  - ordinal: 10
    text: null
    startSeconds: 423
    endSeconds: 433
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=423'
    heldReason: rights
  - ordinal: 11
    text: null
    startSeconds: 433
    endSeconds: 534
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=433'
    heldReason: rights
  - ordinal: 12
    text: null
    startSeconds: 433
    endSeconds: 534
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=433'
    heldReason: rights
  - ordinal: 13
    text: null
    startSeconds: 545
    endSeconds: 612
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=545'
    heldReason: rights
  - ordinal: 14
    text: null
    startSeconds: 612
    endSeconds: 622
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=612'
    heldReason: rights
  - ordinal: 15
    text: null
    startSeconds: 691
    endSeconds: 703
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=691'
    heldReason: rights
  - ordinal: 16
    text: null
    startSeconds: 703
    endSeconds: 746
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=703'
    heldReason: rights
  - ordinal: 17
    text: null
    startSeconds: 746
    endSeconds: 758
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=746'
    heldReason: rights
  - ordinal: 18
    text: null
    startSeconds: 782
    endSeconds: 793
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=782'
    heldReason: rights
  - ordinal: 19
    text: null
    startSeconds: 848
    endSeconds: 858
    watch: 'https://liquidgold.cc/watch/_gVFUEdhCyI?t=848'
    heldReason: rights
---


# Gemma, DeepMind's Family of Open Models — Omar Sanseviero, Google DeepMind

Google DeepMind’s Gemma family is expanding. Join us for a deep dive into the latest models of the Gemma ecosystem. From vibe fine-tuning to Sovereign AI, you'll learn about the latest model capabilities, how to build high-performance applications, and how to get started with open models. Speaker info: - https://x.com/osanseviero - https://www.linkedin.com/in/omarsanseviero/ - https://github.com/osanseviero Timestamps 0:00 Introduction to the Gemma model family 0:41 Evolution from Gemma 3 to Gemma 4 1:21 Overview of the new Gemma 4 capabilities 2:31 Live demonstrations of on-device applications 3:38 LM Arena scores and performance benchmarks 5:07 Apache 2 license transition 5:27 Technical deep dive: The E2B architecture and per-layer embeddings 6:57 Multimodal understanding and multilingual support 8:43 Ecosystem growth and community adoption 10:07 Product integrations, including Android Studio 10:46 Statistics on model downloads and fine-tuning 11:27 Official Gemma variants: Shield Gemma and MedGemma 12:16 Community research and sovereign AI efforts 12:56 Real-world applications, from cancer therapy to offline tasks 14:05 Closing remarks and future outlook --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/\_gVFUEdhCyI-gemma-deepmind-s-family-of-open-models Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney ---

Canonical: https://liquidgold.cc/watch/_gVFUEdhCyI · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md)

> Search chunk text is held until public-source provenance is verified.

## Document chunks

---

#### Chunk 0: Document excerpt (0:15 to 1:22)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 1: Document excerpt (1:22 to 2:38)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 2: Document excerpt (2:38 to 3:44)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 3: Document excerpt (3:44 to 5:10)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 4: Document excerpt (5:10 to 5:32)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 5: Document excerpt (5:32 to 7:03)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 6: Document excerpt (7:03 to 7:13)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 7: Document excerpt (7:13 to 8:54)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 8: Document excerpt (8:54 to 10:12)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 9: Document excerpt (10:12 to 10:57)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 10: Document excerpt (10:57 to 12:26)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 11: Document excerpt (12:26 to 13:02)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 12: Document excerpt (13:02 to 14:08)

> Held: transcript rights do not permit this chunk's text.

---

#### Chunk 13: Document excerpt (14:08 to 15:26)

> Held: transcript rights do not permit this chunk's text.
