/Tech3h ago

Google DeepMind says Gemini Omni Flash leads Video Arena leaderboard and is coming to the developer API soon

Story Overview

Google DeepMind's Gemini Omni Flash has claimed the top Video Arena spot for both text-to-video and image-to-video generation, posting a 158-point Elo jump over Veo 3.1 at 1080p while also leading video-editing preferences. The native multimodal model accepts mixed text, image, audio, and video prompts to create or iteratively refine short clips, with early access already live for paid Gemini app users and YouTube Shorts creators.

1811.5K89212102.9K
Original post
Dumitru Erhan@doomie#162inTech

🚀

Arena.ai@arena

Exciting news: Gemini Omni Flash is now #1 in the Video Arena (both Text-to-Video and Image-to-Video)!

For Text-to-Video this is a massive +158 pt improvement over Veo 3.1 (1080p) and a large +61 pt lead over the next best model, Seedance 2.0.

Congrats @GoogleDeepMind for this huge milestone!

9:48 AM · Jun 11, 2026 · 3.6K Views
Developer Impact

API timing still sits in the weeks-ahead window

Developers will get the model through the Gemini API and Agent Platform soon, but no concrete date, pricing, or rate limits have been shared yet.

Open Question

Editing chops look promising on paper

Conversational multi-turn edits and multimodal references are called out as differentiators, though their edge outside the current human-preference benchmarks remains untested in public workflows.

Sentiment

Positive users congratulate Google on Gemini Omni Flash topping text-to-video and image-to-video arenas for its gains and affordability, while negative users reject the leaderboard results as unreliable and demand real model access instead.

Pos
45.7%
Neg
54.3%
47 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS40.7KBOOKMARKS106LIKES821RETWEETS61REPLIES91
Logan Kilpatrick@OfficialLoganK

Gemini Omni Flash is SOTA at image to video, text to video, and video editing : )

Excited to get this to developers in the API soon!

2hViews 40.7KLikes 821Bookmarks 106
Philipp Schmid@_philschmid

Coming to API soon.

Arena.ai@arena

Exciting news: Gemini Omni Flash is now #1 in the Video Arena (both Text-to-Video and Image-to-Video)!

For Text-to-Video this is a massive +158 pt improvement over Veo 3.1 (1080p) and a large +61 pt lead over the next best model, Seedance 2.0.

Congrats @GoogleDeepMind for this huge milestone!

1hViews 2KLikes 38Bookmarks 6
Logan Kilpatrick@OfficialLoganK

Check out the benchmarks page for Omni: https://deepmind.google/models/gemini-omni/#:~:text=how%20to%20prompt-,Performance,-Gemini%20Omni%20Flash

Logan Kilpatrick@OfficialLoganK

Gemini Omni Flash is SOTA at image to video, text to video, and video editing : )

Excited to get this to developers in the API soon!

2hViews 7.4KLikes 40Bookmarks 6
👩‍💻 Paige Bailey@DynamicWebPaige

📹 Omni is such a strong video generation and editing model, effectively @NanoBanana for video footage!

Coming to the API soon... 👀

Logan Kilpatrick@OfficialLoganK

Gemini Omni Flash is SOTA at image to video, text to video, and video editing : )

Excited to get this to developers in the API soon!

2hViews 2.2KLikes 42Bookmarks 5

> OpenAI trying to shift the responsibility to Deepmind my dawg is washed…

Hensen Juang@basedjensen

@OfficialLoganK @GoogleDeepMind Brah got be real with you folks you guys need to step up and deliver a mythos class model. Side quests like this is irrelevant

2hViews 1.7KLikes 20Bookmarks 2
Tulsee Doshi@tulseedoshi

🏆

Arena.ai@arena

Exciting news: Gemini Omni Flash is now #1 in the Video Arena (both Text-to-Video and Image-to-Video)!

For Text-to-Video this is a massive +158 pt improvement over Veo 3.1 (1080p) and a large +61 pt lead over the next best model, Seedance 2.0.

Congrats @GoogleDeepMind for this huge milestone!

2hViews 511Likes 14Bookmarks 2
Arena.ai@arena

Exciting news: Gemini Omni Flash is now #1 in the Video Arena (both Text-to-Video and Image-to-Video)!

For Text-to-Video this is a massive +158 pt improvement over Veo 3.1 (1080p) and a large +61 pt lead over the next best model, Seedance 2.0.

Congrats @GoogleDeepMind for this huge milestone!

Google DeepMind@GoogleDeepMind

We’re dropping Gemini Omni: our first step towards a model that can create anything from anything - starting with video.

It combines Gemini’s intelligence with our generative media systems - representing a leap forward in world understanding, multimodality, and editing 🧵

3hViews 46.5KLikes 453Bookmarks 87
Arena.ai@arena

Gemini Omni Flash won 82% of its head-to-head battles in Battle Mode (without ties). Incredibly high compared to the next best model’s win-rate of 69%.

3hViews 541Likes 15
Arena.ai@arena

Gemini Omni Flash also ties for #1 in Image-to-Video. Here we see a +77 pt improvement over Veo 3.1.

3hViews 635Likes 9
찡긋@Alignment100

@OfficialLoganK @GoogleDeepMind Google Korea B2C "ME"

2hViews 96Likes 1Bookmarks 1
Squeak Al-Gaib 🐭@SqueakAlGaib

@OfficialLoganK @GoogleDeepMind Seedance users : sure bro.

2hViews 224Likes 11
Hensen Juang@basedjensen

@OfficialLoganK @GoogleDeepMind Brah got be real with you folks you guys need to step up and deliver a mythos class model. Side quests like this is irrelevant

2hViews 284Likes 4
Philipp Schmid@_philschmid

@OfficialLoganK @miketaylorcai @GoogleDeepMind Can’t wait to interact with it

Logan Kilpatrick@OfficialLoganK

Gemini Omni Flash is SOTA at image to video, text to video, and video editing : )

Excited to get this to developers in the API soon!

2hViews 843Likes 3Bookmarks 0
Arena.ai@arena

Check out official Gemini Omni Flash release here: https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-omni/

3hViews 796Likes 3
yves@yvesai0

@OfficialLoganK @GoogleDeepMind Anyone who is using Seedance 2 knows these scores are ridiculous

2hViews 149Likes 6
Nataniel Ruiz@natanielruizg

👾👾👾

Arena.ai@arena

Exciting news: Gemini Omni Flash is now #1 in the Video Arena (both Text-to-Video and Image-to-Video)!

For Text-to-Video this is a massive +158 pt improvement over Veo 3.1 (1080p) and a large +61 pt lead over the next best model, Seedance 2.0.

Congrats @GoogleDeepMind for this huge milestone!

2hViews 252Likes 5Bookmarks 0
Curlheinz@Curlh1

@OfficialLoganK @GoogleDeepMind Yes please! Got any idea on price? Per second or something

2hViews 339Likes 3
Arena.ai@arena

Dive into the Video Arena leaderboard details at: http://arena.ai/leaderboard/text-to-video

3hViews 759Likes 2
ChrisUniverse 🗽@ChrisUniverse

@OfficialLoganK @shlomifruchter @GoogleDeepMind Hey @grok what is HappyHorse and who made it?

2hViews 61
Load more posts