AdGraam News Desk
AI video and image news
The AdGraam News Desk (part of AdGraam, also searched as Adgram or Adgram AI) tracks AI image and video model launches, leaks and rumors from labs, leakers and creators, and checks each claim against its sources. Unconfirmed stories say so, and each page is updated as its story develops.
Articles by AdGraam News Desk

RMBG-3.0: Bria's Open Background Remover, $0.018
Bria RMBG-3.0 is a new background removal model for images and video, priced at $0.018 per image via API. Get the spec table, ad workflow and access details.

Kling Image 4.0 Teased: Coming Soon, What We Know
Kling Image 4.0 was teased by Kling AI on Nov 16, 2026. Get what is confirmed, what is not, how it fits Kling 4.0 video, and an ad workflow breakdown.

Seedance 2.5 Draft Mode: 480p Drafts, 4K Finals
Seedance 2.5 Draft Mode lets you iterate at 480p, then finish at 1080p or 4K. Get the spec table, ad workflow breakdown and access details across platforms.

Vidu S2 Editing: Live Stream Character & Outfit Swaps
Vidu S2 Editing applies character, outfit and style changes to live video. Get the spec table, API cost estimates, ad workflow breakdown and access details.

Omni LoRA Tuning: Google Video Fine-Tuning Leak
Google's Omni LoRA tuning spec reportedly appeared in the Vertex AI API for Gemini Omni video. Get the known details, ad workflow impact and what is unconfirmed.
Vidu S2 Avatar: 720p Live Digital Humans, API Price
Vidu S2 Avatar from ShengShu turns a character image into a live digital human at 720p. Get the spec table, ad workflow breakdown and access details.

NB3 Google Rumor: Gemini Notebook Comic Clue
Google NB3 is rumored after a Gemini Notebook Comic page update, per @testingcatalog. Get the unconfirmed facts, an ad workflow breakdown and what to watch.

Grok Imagine: Image & Video Now Inside Grok
Grok Imagine now lives inside Grok. Image API runs from $0.02 per image. Get the full spec table, ad workflow breakdown and access details for marketers.

MiniMax H3 Ultrafast in Viggle: 10s Video in 10s
MiniMax H3 Ultrafast is live in Viggle: a 10-second 768p clip in about 10 seconds. Get the spec table, ad workflow breakdown and access details here.

P-Video-2-Pro-Reference: Benchmarks & Pricing
P-Video-2-Pro-Reference from Pruna AI ranks #1 on cost among 7 reference-to-video models at ~$0.046 per second. Get the spec table, ad workflow notes and access.

Seedance 2.5 Flash: API Live, $0.152/sec 720p
Dreamina Seedance 2.5 Flash is live on Dreamina and the official API: $0.152/sec at 720p, 30s clips. Get the spec table, ad workflow breakdown and access details.

Recraft V4.1 Flash V2: 1.1s Images, Same Price
Recraft V4.1 Flash V2 is live: 1.1 seconds per image, better text and fine details, same price. Get the spec table, ad workflow breakdown and access details.

Luma Character Swap & Insert: Swap, Add, Remove
Luma AI launched Character Swap & Insert on Nov 11, 2026: swap cast or cars, add a dog, remove people in one shot. Get the ad workflow and access details.

Seedance 2.5 4K Launch: Draft Mode, 10-bit, Access
Dreamina Seedance 2.5 is now in 4K, with 480p Draft previews and 10-bit color. See the spec table, ad workflow breakdown and where you can access it.

Recraft Vectorizer Pro: Cleaner Vectors, All Plans
Recraft Vectorizer Pro is live for every plan, with cleaner curves, straight lines and sharp corners. Get the spec table, ad workflow notes and access details.

PAN 2: 29B Open-Weights World Model, Apache 2.0
PAN 2 from IFM is a 29B open-weights world model, 704x1248 at 24 fps, Apache 2.0. See the spec table, ad workflow breakdown and how to get access.

BouncyBohr: Leaked Google Video Model, 4K Claims
BouncyBohr is a rumored Google video model with 4K and 10-second extensions, per a leaker. Get the spec table, ad workflow breakdown and access details.

Midjourney V8.2 Remove Background: $0.086 per Image
Midjourney V8.2 Remove Background returns one transparent cutout for $0.086 per image in the AdGraam Playground. Get specs, ad workflow tips and access details.

Kling 4.0 Video Edit: 30s Clips & $0.128/sec
Kling 4.0 Video Edit restyles or swaps content in existing 3 to 30 second clips from $0.128/sec. Get the spec table, ad workflow breakdown and access details.

Midjourney V8.2 Image to Video: 4 Clips, 720p
Midjourney V8.2 Image to Video turns one image into four 5-second clips at 480p or 720p. Get the spec table, ad workflow breakdown and access details.

Midjourney V8.2 Text to Image: API, Specs & Pricing
Midjourney V8.2 Text to Image is now callable on a model API: four images per prompt, native 2K HD. Get the spec table, ad workflow breakdown and access details.

Bria Fibo 2 Edit: 4 Image Inputs, $0.04 Per Edit
Bria Fibo 2 Edit takes up to 4 input images and costs $0.04 per image in the AdGraam Playground. Get the spec table, ad workflow breakdown and access details.
HeyGen Avatar V Presenter: 4K Specs & $0.10/sec
HeyGen Avatar V Presenter is live with 720p to 4K output at $0.10 per second in AdGraam Playground. Get the spec table, ad workflow breakdown and access details.

FLUX 3 Image: Native 4K, Klein Open Weights & Price
FLUX 3 Image from Black Forest Labs: native 2K and 4K, bounding boxes, 10 references, $0.041 per image. Get the spec table, ad workflow and access details.

Ideogram V4.5 Text to Image: 2K Specs & $0.03 Price
Ideogram V4.5 Text to Image is live: tiers from $0.03 per image, up to 8 images per run, 6 aspect ratios. Get the spec table, ad workflow and access details.

Krea 3 Edit: 8 Refs, Regional Edit & $0.025/MP
Krea 3 Edit takes up to 8 reference images, supports regional edits and costs $0.025 per megapixel in AdGraam Playground. See specs, ad workflows and access.

Grok Imagine Video Edit: 8.7s Cap, Per-Second Billing
Grok Imagine Video Edit from xAI edits mp4 clips with text instructions, capped at 8.7s and billed per second. Get the spec table, ad workflow and access details.

Midjourney V8.2 Style Transfer: 4 Outputs, $0.129
Midjourney V8.2 Style Transfer restyles an image and returns 4 results for $0.129 per run in AdGraam Playground. Get the spec table, ad workflow and access.

Midjourney V8.2 Image to Image: 4 Outputs, $0.086
Midjourney V8.2 Image to Image is callable on a model API: 4 variations per run, native 2K, $0.086 in AdGraam Playground. Get specs, ad workflows and access.

Midjourney V8.2 Blend: 5 Images In, 4 Out, $0.086
Midjourney V8.2 Blend fuses 2 to 5 images into 4 results at native 2K for $0.086 per run in AdGraam Playground. Get the spec table, ad workflow and access details.

Ideogram V4.5 Edit: 4 Refs, From $0.008/Image
Ideogram V4.5 Edit runs in AdGraam Playground from $0.008 per image with up to 4 references and a high-precision mode. See specs, pricing and ad workflows.

Bria Fibo 2 Text to Image: Specs & $0.04 Price
Bria Fibo 2 (Fibo 2.0) text to image runs at $0.04 per image in AdGraam Playground. Get the spec table, ad workflow breakdown and access details.

MiniMax H3 Image Highspeed: 20MP, $0.0219/Image
MiniMax H3 Image Highspeed text to image is live: sizes up to 20MP, from $0.0219 per image. Get the spec table, ad workflow breakdown and access details.

MiniMax H3 Image Highspeed Edit: 16 Refs, $0.0219
MiniMax H3 Image Highspeed Edit takes up to 16 reference images and outputs up to 20MP from $0.0219 per image. Get the spec table, ad workflow and access.

VEED Dubbing: 86 Languages, $0.11/sec Video Translate
VEED Dubbing Video Translate dubs up to 10 minutes into 86 languages with voice cloning and lip-sync at $0.11/sec. Get specs, ad workflow and access details.

FLUX 3 Image Edit: 10 Refs, 4K & $0.048/MP
FLUX 3 Image Edit from Black Forest Labs takes up to 10 references and outputs up to 4K at $0.048 per megapixel. Get the spec table, ad workflow and access.

MiniMax H3 Image Edit Launch: 16 Refs, $0.009
MiniMax H3 Image Edit is live: up to 16 references, 20 MP output, open weights and $0.009 per 1K/2K edit. Get the spec table, ad workflow tips and access.

VEED Dubbing Pro: 94 Languages, $0.18/sec Video Translate
VEED Dubbing Pro Video Translate dubs 10 minute videos into 94 languages with lip-sync at $0.18 per second. Get the spec table, ad workflow and access details.

Mirelo SFX 2.0: Video to Sound, 10 Min & $0.01/sec
Mirelo SFX 2.0 adds synced sound effects to video, keeps dialogue, runs up to 10 minutes and costs $0.01 per second in AdGraam Playground. Full specs inside.

Seedream 5.5 Layer Decomposition: 99 Layers, Pricing
Seedream 5.5 Layer Decomposition splits an image into a base plus up to 99 PNG layers. Get the spec table, ad workflow breakdown, 3.6 credit pricing and access.

Image Upscale up to 8K: Specs & $0.06 Price
Image Upscale up to 8K is live in AdGraam Playground from 1.5 credits ($0.0075). Get the spec table, ad workflow breakdown and access details here.

Seedream 5.5: 100-Round Edits, RGBA & $0.018/Image
Seedream 5.5 from ByteDance is live: native transparent PNGs, 100-round editing, from $0.018 per image. Get the spec table, ad workflow and access details.

ComfyArt: Open-Source Canvas Wins BFL Challenge
ComfyArt, an open-source ComfyUI design canvas, won Black Forest Labs' challenge. See the spec table, ad workflow breakdown and how to install it.

CreativeField: Free Local ComfyUI Studio Explained
CreativeField is a free, local ComfyUI studio for image, video and 3D, now with a built-in photo editor. Get the spec table, ad workflow notes and access details.

MiniMax-H3-Image 2-Step LoRA: 33B, 4096² Specs
MiniMax-H3-Image gets a 2-step pDMD LoRA: 33.1B model, 4096x4096, 2 steps not 6. See the spec table, ad workflow breakdown and how to access it.

Luma MCP Is Live in ChatGPT and Claude
Luma MCP is live in ChatGPT and Claude, per Luma AI's own post. Get the confirmed details, a spec table, an ad workflow breakdown and access notes.

Video Upscale: 1080P, 2K and 4K in AdGraam Playground
Video Upscale now runs in AdGraam Playground: output 1080P, 2K or 4K while keeping aspect ratio, frame rate and audio. Get specs, ad workflow and access.

MiniMax H3-Image Launch: Open Weights, US Excluded
MiniMax H3-Image is out with open weights, native 4K and 16 references. See the license limits (US excluded), spec table, API pricing and ad workflow tips.

Kling 4.0 Keyframes: 10 Keyframes, 30s Clips
Kling 4.0 Keyframes lets you steer a scene with up to 10 keyframes and 30-second clips. Get the spec table, ad workflow breakdown and access details.

H3Image on Hailuo AI: Spotted Image Model Details
MiniMax H3Image is reportedly live in beta on Hailuo AI. Get what users show, how it relates to MiniMax H3 video (2K, $0.061/sec), and ad workflow notes.

Emma by Nuance Labs: Free Full-Duplex AI Avatar
Nuance Labs reportedly launched Emma, a free video chat AI persona built on its Nuplex model. Get the spec table, ad workflow breakdown and access details.

LTX-2.5: Reported Open Model, 4K HDR & Specs
LTX-2.5 from Lightricks is reported as a downloadable open model with native 4K and HDR. Get the spec table, ad workflow breakdown and access details.

Runway Sightline: 360 Panoramas, Any-Angle Shots
Runway Sightline builds a 360 panorama of any location, then lets you frame shots from any angle. Get the full spec table, ad workflow breakdown and access details.

New Hedra Teased: Launch Expected Nov 10
Hedra teased a new model on 2026-11-09 with a reveal due the next day. Get what is known, the current Avatar free plan details, and an ad workflow breakdown.

Dreamina Marketing Studio Is Live: What It Does
Dreamina Marketing Studio from ByteDance is live. See what is confirmed, the $0.29 new-user offer, plus our ad workflow breakdown and access details.

Higgsfield Drama Ads: 44 Story Cards, 8 Styles
Higgsfield Drama Ads turns a product into a short drama ad with captions and sound. See 8 styles, 44 story cards, the spec table, ad workflow and access.

Seedream 5.5 Is Out: RGBA, 100-Round Editing
Seedream 5.5 (Dreamina Seedream 5.5) is live: native RGBA, 100-round editing, 95% text claim. Get the spec table, ad workflow tips and access details.

OpenAI Tomorrow Teaser: What We Know So Far
OpenAI's Devs account posted a one-word teaser, 'Tomorrow,' on Nov 8, 2026. Get what is confirmed, the rumored 3 models, ad workflow impact and what to watch.

Claude Fable 5.5 Rumor: Leaks, Specs & Status
Claude Fable 5.5 is unannounced. See what the Azure listing, LiteLLM entry and leakers claim, plus Fable 5.1's $10/$50 pricing and what ad teams should do now.

PixVerse FPV Camera Control: Draw a Red Line Path
PixVerse FPV camera control lets you draw a red line to guide a fly-through from one image. Get the spec table, ad workflow breakdown and access details.

Topview T1: 30s 1080p Video, 2.2 Credits/Sec
Topview T1 is a 30-second, 1080p AI video model priced at 2.2 credits per second. Get the spec table, ad workflow breakdown and access details here.

Higgsfield Miracle Preview: Features and Beta Access
Higgsfield Miracle vectorizes video into layers for real-time, pixel-level edits. Here is what Higgsfield claims, the private beta status and MCP access.

Recraft Vary Is Out: Features, Access and Pricing
Recraft added Vary to Recraft Studio on Nov 6, 2026: Vary direction for new compositions, Vary details for small changes, no prompt needed. Access and pricing.

Krea Video Is Out: Features, Pricing and Access
Krea Video launched Nov 5, 2026 with camera controls, 100+ presets, UGC templates, style transfer and sound generation. See features, pricing and access.

P-Video-2-Pro-Replace: Features, Pricing and Access
Pruna AI launches P-Video-2-Pro-Replace, built on MiniMax H3, to swap one or two people in a video with reference images while keeping motion and audio.

Gemini Omni 1.2 Flash Is Out: Features, Pricing and Access
Google's Gemini Omni 1.2 Flash video model is live: 15-second clips, $0.07 per second at 720p, 4K upscaling, multi-turn editing and #1 on LMArena text-to-video.

Viggle Remake Is Out: Open Weights, Pricing and Access
Viggle Remake is live: a video to video agent that recasts characters and edits clips up to 2 minutes, with open weights and pricing from $0.96 per 10 seconds.

Luma Photo Reframe: Features, Access and How It Works
Luma AI launched Photo Reframe on Nov 5, 2026, an image tool that fills the space around a photo to fit new formats. Here is what is known about access.

GPT Image 2.6 Is Out: ChatGPT Images 2.6 Pricing & API
GPT Image 2.6 (ChatGPT Images 2.6) is live from OpenAI with GIFs, SVG export and Flare and Sunburst API models at $30 per 1M image output tokens.

Pyramid-JiT 2B Is Out: Open Weights, Specs and Access
Pyramid-JiT 2B (P-JiT 21K) is Linum's open-weights text-to-image model, trained for about $21K. See specs, Apache 2.0 license, limits and how to run it.

HeyGen Video 1 Is Out: Free Tier & $0.02/sec API
HeyGen Video 1 is live: text-to-video with picture, dialogue and sound in one pass, a free tier, a $0.02 per second API and a new mobile app. Full details.

Kling 4 Flash Is Out: 20s 720p Video & API Specs
Kling 4.0 Flash is now open to everyone on the Kling AI website and via API: 3 to 20 second 720p clips, plugins, and what pricing is known so far.

PixVerse Depth Map Control: Features and How to Use It
PixVerse Depth Map Control lets you recreate the motion of any video with your own character. See how the workflow runs and what it means for ad creatives.

Luma Character Sheet Is Out: Features and Access
Luma Character Sheet turns your reference images into one consistent character sheet. Here is what Luma AI announced, what is unknown, and how to try it.

Pika H3 Flex Is Out: Features, Pricing and Access
Pika H3 Flex is a post-trained MiniMax H3 video model with open weights, 12 references, audio stems and 4K output. See specs, pricing and how to access it.

FLUX 3 Image Klein Is Out: Specs, Benchmarks, Access
FLUX 3 Image Klein (Flux.3 Image Klein) is Black Forest Labs' 5.6B open weights image model. See specs, benchmark ranks, license terms and how to access it.

Topaz Wonder 4 Is Out: MoE Upscaler from $0.043
Topaz Wonder 4 is live in Topaz for Web and Mobile. See the Mixture of Experts upscale, sharpen and denoise model, API pricing from $0.0428 and plan details.

P-H3-Image Open Weights: 8-Step Specs & ComfyUI
Pruna AI open-sourced P-H3-Image, an 8-step image and editing model built from MiniMax-H3. See specs, VRAM needs, ComfyUI access and license terms.

Hy4 Is Out: Tencent Hunyuan Features, Pricing, Access
Tencent launched Hy4 on Nov 4, 2026: a Hunyuan language model with image understanding, computer use and agent skills. Pricing, free quota and access inside.

Krea 3 Image Editing: Native 2K, $0.03 API Price
Krea 3 is Krea's image editing model: native 2K, edits in under 8 seconds, API from $0.03 per image, 50% off on OpenRouter until Nov 10.

Vidu Q4 Preview: 4K, 15 Refs, From $0.014/sec
Vidu Q4 Preview from ShengShu: native 4K, 15 image and 3 audio references, launch price from about $0.014/sec. Get the spec table, ad workflow and access details.

Recraft Logo Maker Public Beta: Free SVG Export
Recraft Logo Maker is live in public beta: answer a few questions, pick a designer-curated style, preview mockups and export an SVG. See features and access.