Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

📢 Big release: New video model 🚀 Introducing Hyperion 2.5: convert 8-bit AI video into true HDR - ProRes 10-bit, EXR 16-bit, H.265 10-bit. AI footage can finally sit alongside live action in professional and VFX pipelines. Available now in Astra, coming soon to Topaz Video

Topaz Labs

31,992 subscribers

343,758 views • 1 month ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

45 Comments

Topaz Labs1 month ago

More details 👇

awesome_visuals1 month ago

wow Monji! excellent news, will try this right away. 💎

Topaz Labs1 month ago

You got a reply from THE Monji himself.

Jason's Vibe Coding1 month ago

converting 8 -bit AI footage into 10-bit ProRes, 16-bit EXR and 10-bit H.265 is a major step toward making generative video usable in professional workflows.

Aldrich Killian1 month ago

Game changer

Farhan1 month ago

i stitch clips from different video models into one piece fairly often and they never match each other tonally, which is half the cleanup. does this help across sources or is it per clip?

Topaz Labs1 month ago

Hey Farhan, Hyperion is an HDR converter, but Starlight Precise 2.6 can help with this a bit when upscaling. If you're still having issues, we recommend scene splitting or processing separately.

ATOM1 month ago

@LudovicCreator Incredible !

Topaz Labs1 month ago

@LudovicCreator

O-Side Media1 month ago

@ATOM_VISUAL @LudovicCreator We should start with upscaling these gifs 😂 but seriously nice upgrade 🎉

Topaz Labs1 month ago

@ATOM_VISUAL @LudovicCreator Done and done.

Pierrick Chevallier | IA1 month ago

Badass news

Topaz Labs1 month ago

With the OG badass

MO1 month ago

WOW, Amazing upgrades. My favorite Upscaling tool 😍😍

Topaz Labs1 month ago

Starlight Precise upscale + Hyperion for HDR = 😍😍😍

MO1 month ago

The Perfect combo ☝️

𝐙𝐞𝐧𝐠 💜1 month ago

Impressive!!

Captain HaHaa1 month ago

Aw yes! Looking forward to testing the limits of this one

Lyra1 month ago

wow. the HDR side looks completely different. like night and day for AI footage. 10 bit ProRes is huge. are you guys testing this for VFX pipelines?

Topaz Labs1 month ago

ABSOLUTELY

Lyra1 month ago

glad you're on board!! have you guys been playing with it yet? curious what the workflow looks like

Nim Eshed 𝕏🦋1 month ago

🫶❤️🫶❤️🫶❤️

The90sPope1 month ago

Fuck yeah, can't wait for it on Topaz Video!

Blend the rules 🇹🇹 🇺🇸1 month ago

I have some criticism, the video on the left look way better, the videos on the right have very obvious clipping with bright saturated elements (we are all viewing sdr video on twitter). should have shown it in log encoding or some other view transform applied to the "raw".

ΛRMIN1 month ago

Great news 😍🔥

Julien | MJM1 month ago

Epic

Umesh1 month ago

wow, great news, can't wait to try!

LudovicCreator1 month ago

Awesome news

Topaz Labs1 month ago

🙌🙌🙌

Will {AI}1 month ago

Pls add local Hyperion 2.5 in Topaz Video

Cloaked Worlds1 month ago

This looks fantastic! Looking forward to testing it with new videos. Will it be possible in the future to upload 10-bit H.265 videos directly into the browser version of Astra? Right now I have to convert my clips to H.264 before I can upload.

Jeff_synthesized1 month ago

Game changer.

𝕄𝔸𝕋ℝ𝕀𝕏1 month ago

This is amazing..

ffswunnd1 month ago

is this the best way to show hdr footage? without tone-mapping and completely blown out? sdr looks way better

Elara AI1 month ago

Hyperion 2.5 converts AI video to true HDR for pro pipelines

Dagmar Bressel1 month ago

Awesome 🙌👏🙌

Evelyn1 month ago

A massive step forward for AI video workflows! Hyperion 2.5 brings professional-grade HDR and VFX quality closer to creators, unlocking new possibilities for visual storytelling.

Claire Bennett1 month ago

A major leap for AI video workflows. Hyperion 2.5 brings impressive HDR quality and professional-grade possibilities to creators.

Eva Morgan1 month ago

Getting AI footage into proper HDR and professional VFX pipelines makes it much more usable beyond social media.

AI Mastery Guide1 month ago

AI footage finally matching live action

Sebastian Buzdugan1 month ago

10-bit output alone cannot recover clipped highlights, so hdr quality depends on reconstruction

HW1 month ago

Will I be able to run locally on a 5090? If so what would a typical 10 seconds of 4k take to process to 16bit?

Topaz Labs1 month ago

Yep, 5090 is good. No firm estimates on time yet. Stay tuned!

John Schoenith1 month ago

Been waiting for this. Let’s party

Evia AI1 month ago

Huge leap toward professional-grade HDR workflows for AI footage

Related Videos

Introducing Runway Ruby. A new model that converts SDR video up to 16-bit HDR in ProRes and EXR sequences. Compatible with any existing uploaded video or generated output up to 30s.

Introducing Runway Ruby. A new model that converts SDR video up to 16-bit HDR in ProRes and EXR sequences. Compatible with any existing uploaded video or generated output up to 30s.

Runway

222,993 views • 1 month ago

Runway Ruby is now available for Max plans and Enterprise. Generate or convert any video into 16-bit EXR sequences or 10 and 12-bit ProRes and HEVC, in BT.2020 color with PQ or HLG. Try it now at the link below.

Runway Ruby is now available for Max plans and Enterprise. Generate or convert any video into 16-bit EXR sequences or 10 and 12-bit ProRes and HEVC, in BT.2020 color with PQ or HLG. Try it now at the link below.

Runway

17,206 views • 1 month ago

LTX HDR beta is now live. Every AI video model before this one output 8-bit SDR only. Fine for social clips. The format falls apart the moment you try to grade. Highlights clip. Shadows crush. AI footage won't composite cleanly against higher-bit-depth CGI. Resolution was never the real issue. Dynamic range was.

LTX HDR beta is now live. Every AI video model before this one output 8-bit SDR only. Fine for social clips. The format falls apart the moment you try to grade. Highlights clip. Shadows crush. AI footage won't composite cleanly against higher-bit-depth CGI. Resolution was never the real issue. Dynamic range was.

LTX

25,137,221 views • 5 months ago

FREE Seedance 2.5 1080p generations – LIVE now. Early access only on Higgsfield. We're giving out FREE 1080p generations for ALL NEW users. The most production-ready AI video model: > High-budget camera work > Native 10-bit color > Lifelike acting & lightning Limited time-offer.

FREE Seedance 2.5 1080p generations – LIVE now. Early access only on Higgsfield. We're giving out FREE 1080p generations for ALL NEW users. The most production-ready AI video model: > High-budget camera work > Native 10-bit color > Lifelike acting & lightning Limited time-offer.

Higgsfield AI 🧩

577,357 views • 1 month ago

1/7: 📢 Big release! Introducing Astra 2 🚀 Our next-gen creative model with a whole new level of controls, including PROMPTS! Try it NOW in Astra app See what it's best for in this short video and 🧵👇

1/7: 📢 Big release! Introducing Astra 2 🚀 Our next-gen creative model with a whole new level of controls, including PROMPTS! Try it NOW in Astra app See what it's best for in this short video and 🧵👇

Topaz Labs

2,626,137 views • 5 months ago

This is literally Fable moment for AI filmmaking. This full movie trailer was created on Buzzy using Seedance 2.5. With Buzzy, creators can: → Generate 30-second cinematic scenes in one go → Create in native 4K with 10-bit color → Use up to 50 image, video, and audio references for stronger character and style consistency This moves AI filmmaking beyond isolated clips. Instead of editing and stitching short individual videos, creators can now build complete stories and cinematic sequences on Buzzy. Seedance 2.5 is coming soon to Buzzy. #Seedance25 #AIFilmmaking #AIVideo

This is literally Fable moment for AI filmmaking. This full movie trailer was created on Buzzy using Seedance 2.5. With Buzzy, creators can: → Generate 30-second cinematic scenes in one go → Create in native 4K with 10-bit color → Use up to 50 image, video, and audio references for stronger character and style consistency This moves AI filmmaking beyond isolated clips. Instead of editing and stitching short individual videos, creators can now build complete stories and cinematic sequences on Buzzy. Seedance 2.5 is coming soon to Buzzy. #Seedance25 #AIFilmmaking #AIVideo

Abhishek

18,091 views • 2 months ago

🎬 Global Launch: Dreamina Seedance 2.5 is now live! Our most cinematic video model yet has arrived, and Dreamina AI is your official platform. Step into a new era of precision control and long-take storytelling: · Native 30s videos in a single generation · A new interactive editing experience: reshape any part of the frame · Long video mode: up to 3 minutes, with consistency that holds throughout · Dreamina AI plugins for Maya and Blender, built for film-grade production pipelines · Up to 50 multimodal references, now with support for 3D white models and green-screen footage — bringing AI generation into real production workflows · More true-to-life lighting and shadows, richer motion detail, and precise timestamp control down to 1 second Dreamina Seedance 2.5 is more than a model. It's an upgrade to what you can create. Now open to Dreamina AI subscribers. 🕐 Available now across Southeast Asia, the Middle East, Africa, Europe, and South America. More countries and regions coming soon. ✨ Watch Aura Design in action with the new model ⏳ Limited time: save up to 32% on credits per generation 🎁 Repost and comment – first 300 users get 1,000 credits in 12 hours #DreaminaSeedance25 #DreaminaAI

🎬 Global Launch: Dreamina Seedance 2.5 is now live! Our most cinematic video model yet has arrived, and Dreamina AI is your official platform. Step into a new era of precision control and long-take storytelling: · Native 30s videos in a single generation · A new interactive editing experience: reshape any part of the frame · Long video mode: up to 3 minutes, with consistency that holds throughout · Dreamina AI plugins for Maya and Blender, built for film-grade production pipelines · Up to 50 multimodal references, now with support for 3D white models and green-screen footage — bringing AI generation into real production workflows · More true-to-life lighting and shadows, richer motion detail, and precise timestamp control down to 1 second Dreamina Seedance 2.5 is more than a model. It's an upgrade to what you can create. Now open to Dreamina AI subscribers. 🕐 Available now across Southeast Asia, the Middle East, Africa, Europe, and South America. More countries and regions coming soon. ✨ Watch Aura Design in action with the new model ⏳ Limited time: save up to 32% on credits per generation 🎁 Repost and comment – first 300 users get 1,000 credits in 12 hours #DreaminaSeedance25 #DreaminaAI

Dreamina AI

641,922 views • 2 months ago

I believe 10 years down the line it would not be uncommon to buy AI intelligence in a new home, for an extra $400-$500k. A centralised intelligence box/rack that every AI harness, including hardware/robots would use. Talking about that future a bit in a video today.

I believe 10 years down the line it would not be uncommon to buy AI intelligence in a new home, for an extra $400-$500k. A centralised intelligence box/rack that every AI harness, including hardware/robots would use. Talking about that future a bit in a video today.

Mehul Mohan

26,112 views • 3 months ago

📢 Big release! Introducing Starlight Precise 2, a new video enhancement model that delivers next-gen REALISM. Say goodbye to plastic & artificial look. Get natural faces, skin textures, and details. Like & comment ASTRA to try now for free. More details in 🧵👇

📢 Big release! Introducing Starlight Precise 2, a new video enhancement model that delivers next-gen REALISM. Say goodbye to plastic & artificial look. Get natural faces, skin textures, and details. Like & comment ASTRA to try now for free. More details in 🧵👇

Topaz Labs

50,345 views • 10 months ago

AI video creates a weird new problem: You can suddenly have too much footage. The generation is fast, but now you have dozens of clips that still need to be reviewed, edited and turned into something coherent. Seedance 2.5 is available on CapCut, and Seedance 2.5 1080p is now live. #CapCutPC #CapCutSeedance25

AI video creates a weird new problem: You can suddenly have too much footage. The generation is fast, but now you have dozens of clips that still need to be reviewed, edited and turned into something coherent. Seedance 2.5 is available on CapCut, and Seedance 2.5 1080p is now live. #CapCutPC #CapCutSeedance25

SARAH

201,094 views • 1 month ago

Relight your videos with SwitchX2.0 🔥 Check out Beeble: Beeble’s AI relighting now runs at 4K, up to 20 seconds, in 10-bit colour, which puts it firmly in production territory rather than test-clip territory. You can reshape the lighting of a scene after it’s filmed and hold onto every bit of detail in the highlights and shadows. Built for filmmakers, editors, and advertisers who need delivery-ready footage. It’s the most production ready AI tool out there imo! Our mission is to create 100,000 new jobs for creatives, looking to learn and stay ahead in AI? Join GenHQ: #BeebleAI #AIRelighting #4KVideo #AIVideoEditing #creativeai

Relight your videos with SwitchX2.0 🔥 Check out Beeble: Beeble’s AI relighting now runs at 4K, up to 20 seconds, in 10-bit colour, which puts it firmly in production territory rather than test-clip territory. You can reshape the lighting of a scene after it’s filmed and hold onto every bit of detail in the highlights and shadows. Built for filmmakers, editors, and advertisers who need delivery-ready footage. It’s the most production ready AI tool out there imo! Our mission is to create 100,000 new jobs for creatives, looking to learn and stay ahead in AI? Join GenHQ: #BeebleAI #AIRelighting #4KVideo #AIVideoEditing #creativeai

Rourke Heath

27,509 views • 20 days ago

Beeble has released #SwitchLight 3.0, the new version of its video-to-PBR AI model It extracts #PBR texture maps from video, making it possible to relight live footage - and is now available for use offline, as well as in Beeble's browser app Beeble

Beeble has released #SwitchLight 3.0, the new version of its video-to-PBR AI model It extracts #PBR texture maps from video, making it possible to relight live footage - and is now available for use offline, as well as in Beeble's browser app Beeble

CG Channel

92,955 views • 11 months ago

Seedance 2.5 First Test Something feels a little off but I'm not entirely sure what it is. Seedance 2.0 wasn't like this. It seems to morph a bit during fast action scenes. It’s a really expensive model to test, so we'll see how many attempts we can make. Being able to generate a full 30-second video in one go is nice. In this video, one of the guns disappears and then comes back.

Seedance 2.5 First Test Something feels a little off but I'm not entirely sure what it is. Seedance 2.0 wasn't like this. It seems to morph a bit during fast action scenes. It’s a really expensive model to test, so we'll see how many attempts we can make. Being able to generate a full 30-second video in one go is nice. In this video, one of the guns disappears and then comes back.

Kōda

41,106 views • 2 months ago

I asked Claude Opus 5.5 to make an animated video about a robot who realises he's stuck in an AI-generated world. I wanted to really push it: 12 worlds, 12 styles, from 8-bit to claymation to pencil sketch, you can really see how well this model has imagined all worlds. This is the best animated video you will see today made by Opus 5.5, I promise.

I asked Claude Opus 5.5 to make an animated video about a robot who realises he's stuck in an AI-generated world. I wanted to really push it: 12 worlds, 12 styles, from 8-bit to claymation to pencil sketch, you can really see how well this model has imagined all worlds. This is the best animated video you will see today made by Opus 5.5, I promise.

Pradeep Kapoor

64,113 views • 10 days ago

I’ve always been a huge anime fan, especially Chainsaw Man. So I wanted to push my limits a bit and try creating a live-action version of Denji vs. Katana Man. Since anime fight scenes are known for their insane speed and intensity, translating that energy into live action became a challenge I genuinely enjoyed. This whole sequence was built using an image-to-video workflow with MiniMax Design (H3) 02 and Kling AI 2.5 Turbo, with the base images from Nano Banana Pro on Freepik (now Magnific) . Edited in CapCut Pro and upscaled with Topaz Labs for that final punch. Such a wild experiment, but honestly super fun to make. What do you think about the result?

I’ve always been a huge anime fan, especially Chainsaw Man. So I wanted to push my limits a bit and try creating a live-action version of Denji vs. Katana Man. Since anime fight scenes are known for their insane speed and intensity, translating that energy into live action became a challenge I genuinely enjoyed. This whole sequence was built using an image-to-video workflow with MiniMax Design (H3) 02 and Kling AI 2.5 Turbo, with the base images from Nano Banana Pro on Freepik (now Magnific) . Edited in CapCut Pro and upscaled with Topaz Labs for that final punch. Such a wild experiment, but honestly super fun to make. What do you think about the result?

Dinda Prasetyo

211,083 views • 10 months ago

If you're interested in learning Rive but aren't sure where to start, I'm excited to bring you a new short video that will get you up and running in about 10 minutes. You'll create an interactive design using Rive's vector shape tools, inputs, and even use a little bit of data binding. This is for true beginners, so please dive in and enjoy the video! I can't wait to see what you create in Rive.

If you're interested in learning Rive but aren't sure where to start, I'm excited to bring you a new short video that will get you up and running in about 10 minutes. You'll create an interactive design using Rive's vector shape tools, inputs, and even use a little bit of data binding. This is for true beginners, so please dive in and enjoy the video! I can't wait to see what you create in Rive.

Joey Judkins

12,547 views • 8 months ago

introducing lipsync-1.9-beta, setting a new standard for lipsync quality it’s zero-shot—no training data needed. generate and edit natural speech seamlessly in any live action, animation, and AI-generated video this is the biggest update we’ve ever released, it’s the most natural lipsyncing model in the world 🔥 available now, try it in our app or via API 👇

introducing lipsync-1.9-beta, setting a new standard for lipsync quality it’s zero-shot—no training data needed. generate and edit natural speech seamlessly in any live action, animation, and AI-generated video this is the biggest update we’ve ever released, it’s the most natural lipsyncing model in the world 🔥 available now, try it in our app or via API 👇

sync. labs

129,213 views • 1 year ago

Kling AI Cannes Showcase — House of David & Trailer of The Old Stories: Moses Deep integration into Hollywood-grade pipelines. House of David, the first Hollywood production to openly discuss the use of AI video generation technology in the production process on an industrial level, has captivated over 44M viewers worldwide, ranked among the U.S.’s top 10 new series debuts, and reached #1 on Prime Video in the U.S. Kling AI works seamlessly with live-action production, delivering consistent, high-quality shots at scale and meeting strict industrial delivery standards — proving AI as a reliable, efficient new foundation for global filmmaking.

Kling AI Cannes Showcase — House of David & Trailer of The Old Stories: Moses Deep integration into Hollywood-grade pipelines. House of David, the first Hollywood production to openly discuss the use of AI video generation technology in the production process on an industrial level, has captivated over 44M viewers worldwide, ranked among the U.S.’s top 10 new series debuts, and reached #1 on Prime Video in the U.S. Kling AI works seamlessly with live-action production, delivering consistent, high-quality shots at scale and meeting strict industrial delivery standards — proving AI as a reliable, efficient new foundation for global filmmaking.

Kling AI

402,183 views • 4 months ago

Have you used quantization with an open source machine learning library, and wondered how quantization works? How can you preserve model accuracy as you compress from 32 bits to 16, 8, or even 2 bits? In our new short course, Quantization in Depth, taught by Hugging Face's Marc Sun and @younesbelkada, you'll learn to implement variants of linear quantization, such as asymmetric and symmetric modes, from scratch. You'll also quantize at different granularities (per-tensor, per-channel, per-group) to maintain performance. You’ll then construct a quantizer to compress any open source deep learning model’s dense layers to 8-bit precision. Finally, you’ll practice quantizing weights into 2 bits by packing four 2-bit weights into a single 8-bit integer. If you've ever run a large open source model on your laptop, you've likely benefited from someone's work in quantization. Come learn how this key technique works under the hood! Please sign up here:

Have you used quantization with an open source machine learning library, and wondered how quantization works? How can you preserve model accuracy as you compress from 32 bits to 16, 8, or even 2 bits? In our new short course, Quantization in Depth, taught by Hugging Face's Marc Sun and @younesbelkada, you'll learn to implement variants of linear quantization, such as asymmetric and symmetric modes, from scratch. You'll also quantize at different granularities (per-tensor, per-channel, per-group) to maintain performance. You’ll then construct a quantizer to compress any open source deep learning model’s dense layers to 8-bit precision. Finally, you’ll practice quantizing weights into 2 bits by packing four 2-bit weights into a single 8-bit integer. If you've ever run a large open source model on your laptop, you've likely benefited from someone's work in quantization. Come learn how this key technique works under the hood! Please sign up here:

Andrew Ng

198,671 views • 2 years ago