New AI Models on Pixraft (November 2026): Kling 4.0, GPT Image 2.6, Seedream 5.5, Gemini Omni 1.2 Flash and More
Kling 4.0, GPT Image 2.6, Seedream 5.5, Gemini Omni 1.2 Flash, Vidu Q4, PixVerse V6 Max, Krea V3 and more are live in Pixraft. What each is for.

Pixraft's catalog now includes Kling 4.0, Gemini Omni 1.2 Flash, Vidu Q4, PixVerse V6 Max, GPT Image 2.6, Seedream 5.5, Krea V3, OmniHuman 2.0, and more. The November 11 update adds 64 models: 21 for images, 33 for video, 3 talking-avatar models, and 7 for audio. Earlier versions such as GPT Image 2.5 and Seedream 5.0 stay in the catalog. Models the provider discontinued, such as the Vidu Q4 Preview, were removed: 55 retired models are gone from the catalog, and the model catalog lists what is live.
This post covers only what is new since the October 2026 update. Every model uses the same credit balance, and the exact credit cost is shown before each generation. Specs come from each model's WaveSpeed provider page; the main ones are linked at the bottom.
What's new in the November 11 update?
| Model | Maker | What it's for | Where in Pixraft |
|---|---|---|---|
| Kling 4.0 | Kuaishou | Clips of 3 to 30 seconds with stereo audio, plus a Flash tier at 720p | Video Studio |
| Gemini Omni 1.2 Flash | Short videos with synchronized audio, 3 to 15 seconds, up to 4K | Video Studio | |
| Vidu Q4 | ShengShu Technology | Image-to-video and multi-reference video from 540p up to 4K | Video Studio |
| PixVerse V6 Max | PixVerse | 1 to 15 second clips from 360p to 1080p, with optional audio | Video Studio |
| GPT Image 2.6 | OpenAI | Flare for everyday speed, Sunburst for fine detail, up to 4K | Image Studio |
| Seedream 5.5 | ByteDance | Native transparent backgrounds at 1K to 2K, edits from up to 10 references | Image Studio |
| Krea V3 | Krea | Quality and speed variants, plus edits from up to eight references | Image Studio |
| OmniHuman 2.0 | ByteDance | A talking-avatar video from an image and an audio recording | Lip Sync Studio |
| ElevenLabs Sound Effects v3 | ElevenLabs | Sound effects, ambience, and multi-part audio scenes from text | Audio Studio |
The sections below add the rest of the update: MiniMax H3 Lite and Highspeed, HeyGen Video 1.0, FLUX.3 Klein, MiniMax H3 Image, HeyGen Avatar IV, and Mirelo SFX 2.0.
Which new AI video models are live in Pixraft?
Thirty-three video models are new in Video Studio, across text-to-video, image-to-video, reference-to-video, and video edit.
| Model | Maker | What it's for | Where in Pixraft |
|---|---|---|---|
| Kling 4.0 and 4.0 Flash | Kuaishou | Text, image, and reference video, plus prompt-based video edit | Video Studio |
| Gemini Omni 1.2 Flash | Text, image, and reference video, plus natural-language video edit | Video Studio | |
| Vidu Q4 | ShengShu Technology | Image-to-video and multi-reference video, with optional audio | Video Studio |
| PixVerse V6 Max | PixVerse | Text and image to video, with native audio and multi-clip generation | Video Studio |
| MiniMax H3 Lite and Highspeed | MiniMax | 4 to 15 second videos at 768p, from text, an image, or references | Video Studio |
| HeyGen Video 1.0 | HeyGen | Short videos from text, a first frame, or image and video references | Video Studio |
| Seedance 2.0 and 2.5 reference-to-video | ByteDance | Clips guided by reference images, videos, and audio, with Turbo variants | Video Studio |
Kling 4.0 and Kling 4.0 Flash
Kling 4.0 text-to-video generates 3 to 30 second clips at 720p or 1080p with synchronized stereo audio, in 16:9, 9:16, 1:1, or 21:9. Image-to-video animates a first frame with an optional last frame. Reference-to-video follows up to 10 reference images, 5 reference videos, and 3 reference audios, and video edit rewrites an existing 3 to 30 second clip from a text instruction while keeping the scene, subject, and motion.
Kling 4.0 Flash is described as the fast, lower-cost tier for quick iteration. It outputs 720p clips of 3 to 20 seconds.
Gemini Omni 1.2 Flash
Gemini Omni 1.2 Flash generates clips of 3 to 15 seconds in 16:9 or 9:16, at resolutions from 360p up to 4K, with synchronized audio. Image-to-video accepts a last frame, reference-to-video takes up to 10 images and 3 reference videos, and video edit applies natural-language changes to an existing clip.
The rest of the video lineup
Vidu Q4 image-to-video runs from 540p up to 4K, and its reference-to-video mode accepts up to 15 images and 3 audio references. PixVerse V6 Max produces 1 to 15 second clips from 360p to 1080p, with optional native audio and multi-clip generation. MiniMax H3 Lite and Highspeed generate 4 to 15 second videos at 768p, and Lite also offers 480p.
Seedance 2.0 and 2.5 now have reference-to-video modes, with Fast, Mini, and Turbo variants for Seedance 2.0 and a Turbo variant for 2.5. Each accepts reference images, reference videos, and reference audio next to the prompt.
Which new AI image models are live in Pixraft?
Twenty-one image models are new in Image Studio: eleven for text-to-image and ten for editing with reference images.
| Model | Maker | What it's for | Where in Pixraft |
|---|---|---|---|
| GPT Image 2.6 | OpenAI | Flare and Sunburst tiers with five quality levels up to 4K, plus editing | Image Studio |
| Seedream 5.5 | ByteDance | Native transparent backgrounds at 1K, 1.5K, and 2K, plus multi-reference edits | Image Studio |
| Krea V3 | Krea | Quality and speed variants with seed control, plus edits from up to eight references | Image Studio |
| FLUX.3 Klein | Black Forest Labs | Base and Distilled variants at 1K to 2K, with edits from up to three references | Image Studio |
| MiniMax H3 Image | MiniMax | Photorealistic images at 1K up to 4K, with multi-reference editing | Image Studio |
| Bria FIBO 2.0 | Bria | Text-to-image from a prompt with an optional reference image, plus an edit model with optional masks | Image Studio |
GPT Image 2.6 splits into Flare, the fast everyday tier, and Sunburst, the precision tier that takes longer per image for steadier details across repeated edits. Seedream 5.5 can output a transparent background natively, which suits stickers, product cut-outs, and layered design. For a use-case comparison across image models, read Best AI image generators in 2026.
Which talking-avatar models are new?
Three talking-avatar models are new in Lip Sync Studio. Each turns a portrait and a speech recording into a video.
| Model | Maker | What it's for | Where in Pixraft |
|---|---|---|---|
| OmniHuman 2.0 | ByteDance | A lifelike talking-avatar video from an image and an audio recording | Lip Sync Studio |
| HeyGen Avatar IV | HeyGen | A talking video from a portrait and speech, with optional performance directions | Lip Sync Studio |
| P-Video 2 Pro Avatar | Pruna AI | A talking portrait video at 480p or 768p | Lip Sync Studio |
Use only portraits you have the right to animate. A voice-over from Audio Studio can go straight into these models, and the talking avatar guide walks through the workflow.
Which new AI audio models are live in Pixraft?
Seven audio models are new in Audio Studio: sound effects, music, and speech.
| Model | Maker | What it's for | Where in Pixraft |
|---|---|---|---|
| ElevenLabs Sound Effects v3 | ElevenLabs | Sound effects, ambience, and multi-part audio scenes from text | Audio Studio |
| Mirelo SFX 2.0 | Mirelo | Sound effects from text, or synchronized to an existing video | Audio Studio |
| Sonilo Video-to-Music | Sonilo | Background music synced to a video's mood, pacing, and scene transitions | Audio Studio |
| YuE2 3B | Open-source (YuE) | A new musical rendition from an audio reference, lyrics, and a style description | Audio Studio |
| HeyGen Voice | HeyGen | Text to speech with a choice of public voices, plus a Turbo tier | Audio Studio |
How is pricing shown for the new models?
Every generation shows its exact credit cost on the Generate button before you commit. Most new models are priced from the settings you pick, such as duration, resolution, and quality, so the cost updates as you change them.
Some models are billed by the length of a file you provide. That covers video edit, talking avatars, video-to-music, and reference-to-video with reference clips. For these, the cost appears once the file is attached, because the price depends on how long the clip is. Video edit, talking-avatar, and video-to-music jobs are capped in length. Reference-to-video is priced from the clips you attach, so check the cost before you generate. The one exception is HeyGen Video 1.0: its reference clips over 15 seconds are refused before any charge, because only the first 5 seconds of each are used. Big video jobs also cost fewer credits since Nov 15, 2026; the pricing post has examples.
Which new model should you start with?
Start from the job, not the model name:
- A long single clip with sound: Kling 4.0, up to 30 seconds.
- Fast, lower-cost drafts of video ideas: Kling 4.0 Flash or MiniMax H3 Highspeed.
- Editing footage you already have: Kling 4.0 Video Edit or Gemini Omni 1.2 Flash Video Edit.
- Keeping a subject consistent across shots: reference-to-video on Kling 4.0, Vidu Q4, or Seedance 2.5.
- Transparent-background assets: Seedream 5.5.
- Fine detail and repeated edits: GPT Image 2.6 Sunburst.
- A talking presenter from one photo: OmniHuman 2.0 or HeyGen Avatar IV.
- Sound for a clip you already have: Mirelo SFX 2.0 or Sonilo Video-to-Music.
For video model trade-offs in more depth, the Kling 3.0 vs Veo 3.1 vs Seedance 2.5 vs Wan 3.0 comparison goes through each use case.
How do you try a new model in Pixraft?
- Open Video Studio, Image Studio, or Audio Studio.
- Search the model picker for the model name and choose the variant that matches your input: text, an image, references, or an existing clip.
- Set the options the model offers, such as duration, resolution, aspect ratio, and audio. The studio only shows the settings the selected model supports.
- Check the credit cost on the Generate button. Images start at 2 credits; video and audio costs depend on the model and settings such as length and resolution.
- Generate a short test before a long or high-resolution version, and save the prompt and settings of anything you may want to reproduce.
What does "now live" mean?
It means the model is in Pixraft's catalog and can be selected in the studio today. It is not a quality ranking, and it does not mean every model fits every brief. Provider schemas and limits can change, so the model pages and the live form in the studio are the source of truth.
Failed generations, and outputs blocked by automated content moderation, are refunded to your balance automatically. Sora is not part of the lineup: OpenAI shut down the Sora API on September 24, 2026, and the Sora alternative page covers what to use instead.
New accounts get 10 free starter credits for Fast Image previews after signup, with no card required; a credit pack or plan unlocks the new models.
Questions answered
- What new AI models are available on Pixraft in November 2026?
- For video: Kling 4.0 and Kling 4.0 Flash, Gemini Omni 1.2 Flash, Vidu Q4, PixVerse V6 Max, MiniMax H3 Lite and Highspeed, HeyGen Video 1.0, and reference-to-video for Seedance 2.0 and 2.5. For images: GPT Image 2.6, Seedream 5.5, Krea V3, FLUX.3 Klein, MiniMax H3 Image, and Bria FIBO 2.0. For talking avatars: OmniHuman 2.0 and HeyGen Avatar IV. For audio: ElevenLabs Sound Effects v3 and Mirelo SFX 2.0.
- What can Kling 4.0 do in Pixraft?
- According to its model documentation, Kling 4.0 generates clips of 3 to 30 seconds at 720p or 1080p with synchronized stereo audio, in four aspect ratios including 21:9. It runs text-to-video, image-to-video, reference-to-video, and video edit. Kling 4.0 Flash is the faster, lower-cost tier at 720p, with clips of 3 to 20 seconds.
- Are GPT Image 2.5 and Seedream 5.0 still available?
- Yes. Earlier versions stay in the catalog next to the new ones, so existing workflows keep running. GPT Image 2.6 adds Flare and Sunburst tiers, and Seedream 5.5 adds native transparent-background output. Pick the version in the model picker; the credit cost for each is shown before you generate.
- How is pricing shown for the new video and avatar models?
- Every generation shows its exact credit cost on the Generate button before you commit. Models billed by the length of a file you provide, such as video edit, talking avatars, video-to-music, and reference-to-video with reference clips, show their cost once the file is attached, because the price depends on the clip length.
- Do I need a subscription to try the new models?
- No. New accounts get 10 free starter credits for Fast Image previews after signup, with no card required. A credit pack or plan unlocks the new models, and failed generations and outputs blocked by automated content moderation are refunded to your balance automatically.
Sources
Related articles

Industry Insights
How Much Does a Long or 4K AI Video Cost? Pixraft Credit Prices (Nov 15, 2026)

Industry Insights
Pixraft Developer API, MCP Server and CLI: Run AI Image, Video and Audio Models From Code or Claude

Industry Insights
New AI Models on Pixraft (October 2026): Nano Banana 2.1, GPT Image 2.5, MiniMax H3 Singularity, Kling 3.0, Veo 3.1 and More
Explore Pixraft Multi-Model Studio
Generate images with Nano Banana Pro and FLUX, video with Kling 4.0, Veo 3.1 & Seedance 2.5, and lip-synced audio in one workspace.
Create Free Account →