AI Image and Video Release Notes
Release notes for AI image, text-to-video, image editing, and generative media platforms
Products (14)
Latest AI Image and Video Updates
- Oct 13, 2026
- Date parsed from source:Oct 13, 2026
- First seen by Releasebot:Oct 14, 2026
Kling on Runway Dev
Runway API adds Kling video generation to its developer platform, with Kling 2.5 Turbo Pro, Kling 3.0, and Kling O3 models supporting text, image, and video inputs, multiple formats, audio options, and configurable durations.
October 13th, 2026 - Kling is now available on Runway Dev.
Kling 2.5 Turbo Pro
Kling 2.5 Turbo Pro (kling2.5_turbo_pro) generates from text or a start frame, with an optional last frame, at 5 or 10 seconds (default 5). Text-to-video ratio is 1920:1080 (the default), 1440:1440, or 1080:1920 (1280:720, 720:720, and 720:1280 are still accepted). Image-to-video follows the input image. Billed at 12 credits per second.
Kling 3.0
Kling 3.0 Standard (kling3.0_standard), Kling 3.0 Pro (kling3.0_pro), and Kling 3.0 4K (kling3.0_4k) generate from text or first and last frames for 3–15 seconds (default 5), with optional audio (audio defaults to true) and multishot prompts via providerSettings.multiPrompt (up to 6 segments, 3–15 seconds total).
Text-to-video ratio is 1280:720, 720:1280, or 960:960 (Standard), 1920:1080, 1080:1920, or 1440:1440 (Pro), and 3840:2160, 2160:3840, or 2880:2880 (4K).
Billed at 9 credits per second without audio and 13 with audio (Standard), 12 and 17 (Pro), and 43 either way (4K).
Kling O3
Kling O3 Standard (klingO3_standard), Kling O3 Pro (klingO3_pro), and Kling O3 4K (klingO3_4k) use those same ratios and durations, and also accept up to 7 reference images on image-to-video (keyframes count toward that total) and video-to-video edits of a clip up to 10 seconds.
O3 is billed at 13, 17, and 43 credits per second, including when audio is false. Kling O3 video-to-video is billed per second of the input video.
Additional details
cfgScale (0–1, default 0.5) is available on Kling 2.5 Turbo Pro and Kling 3.0.
Use the text to video, image to video, and video to video endpoints to get started.
Original source - Oct 12, 2026
- Date parsed from source:Oct 12, 2026
- First seen by Releasebot:Oct 14, 2026
Ideogram 4.5 on Runway Dev
Runway API adds Ideogram 4.5 to Runway Dev, enabling image generation with legible text and layout, reference images, and precise editing with masks. Users can choose image ratios, output counts, and quality levels through the text-to-image endpoint.
October 12th, 2026 - Ideogram 4.5 (ideogram_4_5) is now available on Runway Dev.
Generate images with legible typography and layout from a prompt of up to 10,000 characters, with up to 5 reference images and outputCount from 1–8. ratio is a width:height size across 1K and 2K (for example 1024:1024, 2560:1440, 1440:2880), or auto (the default) to let the model choose. Set preciseEdit: true to edit the first reference image while preserving its composition and dimensions, or pass a black-and-white maskUri to edit only the black areas. quality is very_low, low, medium, or high, billed at 1, 4, 8, and 25 credits per image, multiplied by outputCount; it defaults to medium with reference images and high otherwise.
Use the text to image endpoint to get started.
Original source All of your release notes in one feed
Join Releasebot and get updates from Runway AI and hundreds of other software products.
- Oct 12, 2026
- Date parsed from source:Oct 12, 2026
- First seen by Releasebot:Oct 14, 2026
Ideogram 4.0 on Runway Dev
Runway API adds Ideogram 4.0 image generation, with legible typography, reference-image remixing, multiple sizes, and adjustable quality.
October 12th, 2026 - Ideogram 4.0 (ideogram_4_0) is now available on Runway Dev.
Generate images with legible typography from a prompt of up to 4,000 characters, or remix a single reference image, with outputCount from 1–4. ratio is one of nine 2K width:height sizes (2048:2048 by default, plus 2560:1440, 1440:2560, 2304:1728, 1728:2304, 2496:1664, 1664:2496, 3072:1280, 1280:3072). quality is low, medium (the default), or high, billed at 4, 7, and 11 credits per image, multiplied by outputCount.
Use the text to image endpoint to get started.
Original source - Oct 9, 2026
- Date parsed from source:Oct 9, 2026
- First seen by Releasebot:Oct 11, 2026
Weekly Update: October 3 to October 9, 2026
Civitai adds Crucibles community competitions, new generation options like Nano Banana 2.1, MiniMax H3 Max, HeyGen, Vidu Q4 and SDPose, plus creator tools for tips, announcements and score insights. It also delivers quality of life updates, API improvements and bug fixes.
Civitai Changelog: October 3 to October 9, 2026
🚀 New Features
Crucibles
- Crucibles are open to everyone. Enter, and the community votes on entries two at a time until the best ones rise. Rankings stay hidden until it ends, and the prize pool is the host's seed plus every entry fee. Anyone with a creator score of 5,000 can run one. The launch article has the full rundown.
- Since launch: crucibles are listed on both civitai.com and civitai.red, and you pay the entry fee in your site's Buzz. Hosts can require entries made with certain base models, and can warn late entrants or close entries before the end. You can follow a crucible for a reminder before it ends, and remove your own entry while it runs (the fee stays in the pool).
- Judging got a redesign, with a progress count and streak, a briefing on your first visit, and a mark on each crucible card showing whether you have pairs left to judge. Judging needs a creator score of 500.
- Each creator wins at most one prize per crucible, so the next creator moves up. Your stats show your average finish instead of a win rate, which punished anyone who entered big crucibles.
Generation
- Nano Banana 2.1, with 1K, 2K and 4K output and extra-wide and extra-tall ratios up to 8:1 and 1:8.
- MiniMax H3 Max and HeyGen join MiniMax H3 as new versions. Max has a Turbo toggle and goes up to 1080p.
- Vidu Q4 joins Vidu, for image to video and for video from up to 15 references, at up to 4K.
- SDPose is a new pose ControlNet option, and it picks up stylized and anime characters.
Creators
- Models with a licensing fee can now receive generation tips, and tips have their own tab on the creator dashboard and their own column in Creator Studio, instead of being counted as Compensation.
- Creator announcements can carry up to three link buttons for members. Line breaks you type now show, and off-site buttons open a "you're leaving Civitai" page first.
✨ Quality of Life
- Updated Terms of Service on civitai.com and civitai.red. Violence against minors, or a violent setting where harm to a minor is shown or implied, is now named as prohibited. Ordinary play and minor accidental injuries aren't covered. On civitai.com, the broad "dangerous, harmful or inappropriate situations" line is replaced by that item and one prohibiting sexual content involving minors in any style. You'll be asked to accept the Terms again on both sites.
- The Creator Score card on your account explains how the score is earned.
- Daily article limits follow your total Creator Score, like every other score unlock, rather than a separate articles score.
- The Legends leaderboard is now called Legacy, and the Legend referral rank is now Ambassador.
- Pages jump around less while loading, on model pages, post pages and the home feed.
- The generator shows its prompt hints again, like "Your prompt goes here...".
- Reset in the generator clears only the model family you're on, instead of sending you back to the default model.
- Signed out, adding an image in the generator asks you to sign in instead of attempting an upload that fails.
- If your phone's photo picker hands over a file that can't be read, you're offered "Choose from Files" instead of an error.
- The New & Upcoming Creators home block spreads across more creators. One creator could fill a third of it.
- Stickers have a Hide controls toggle, so you can see how one will look before you place it.
- The public API has a post endpoint with its title, description and images, video thumbnails on image results, and a post's images in post order.
🐛 Bug Fixes
- The generator's queue counter could fill up and never clear after one failed job, so generating stopped until you refreshed.
- Hunyuan video was failing with "not enabled for generation" from around September 24.
- Applying Hires Fix or Face Fix changed your Denoise, and switching back to another model family lost the settings you'd chosen there.
- Image To Image from Extract Metadata crashed the page.
- The cost box and source image uploads could spin forever. A failed upload now shows an error, and an image your device can't read says so.
- Uploads on a phone could sit at their last progress forever when the connection dropped without closing.
- Signed-out visitors following a link from civitai.com to civitai.red were sent to a login page instead of the content.
- Crucible entries: an entry whose image was deleted still used up your entry slot, Skip could keep serving the same pair in small crucibles, and you were sometimes paired against two entries from the same creator.
- A bounty supporter could award the bounty to their own entry, and a bounty owner could enter their own bounty.
- The Hide model button showed on your own models and could vanish on someone else's.
- Creator Studio's Tips card counted credits that weren't tips, like support refunds.
- The collection invite link's copy button could copy nothing and still say "Copied".
- The Blue Buzz rewards link on the Buzz Dashboard did nothing on the Blue tab.
🧪 In Testing
Available to a limited group while we finish testing. If you can't see these yet, that's expected.
- Creator Journey. Your Creator Score tier with its badge, what each tier unlocks, achievements for what you've made and earned, and a Creator Showcase of the top creators.
- Qwen Image 2.1 generation and editing in the generator.
🌟 Coming Soon
- Creator Program banking change. From November 1, generation compensation stops counting toward what you can bank. It's still paid to you as it is now.
📣 Official Posts
- Crucibles: Let the Community Pick the Winner - October 2
- H3 Content Contest: Six Seconds to Win Somebody Over - October 2
- Oct 8, 2026
- Date parsed from source:Oct 8, 2026
- First seen by Releasebot:Oct 10, 2026
Testing Thinking Mode
Midjourney tests a new Thinking Mode for image generation on its Alpha website, letting users rerun jobs with added reasoning to improve prompt accuracy, typography, and coherence.
hi everyone! We're testing a new "Thinking Mode" for image generation on our Alpha website (alpha.midjourney.com) and we'd love your feedback.
To run it go under the lightbox for a job and click "Rerun (Thinking)" in the lower right corner
We're finding in tests that this helps make the model better at prompt accuracy, typography, and coherence.
We'd love your help testing it on your images and telling us where it does/doesnt help and what other sort of issues you might find.
In the future we might make this available broadly as a 'thinking' mode or as a way to 'add more thinking' after the job.
We're trying to both improve performance and just generally investigate this direction and would love your thoughts.
Please drop us examples and opinions in #ideas-and-features
Thank you so much! ❤️
Original source - Oct 8, 2026
- Date parsed from source:Oct 8, 2026
- First seen by Releasebot:Oct 10, 2026
Alpha Changelog - 10/7/26
Midjourney adds shared folders and Thinking Mode for rerunning 8.2 jobs with smarter prompt correction, while also improving organization, moodboards, video prompts, and a range of editor, submit, and account fixes across alpha.midjourney.com.
Hi everyone, here's this week's round of changes on alpha.midjourney.com. The highlight this week is shared folders, a new way to collaborate on Midjourney. Many of these improvements came straight from #alpha-ideas-and-bugs and office hours, so keep your thoughts coming!
THINKING MODE
You can now rerun jobs made with 8.2 (standard and edit) with "thinking.” Midjourney looks at your image, finds where it missed your prompt (wrong objects, layout, anatomy, text) and generates again. This is an early version so we'd love to hear your feedback.
SHARED FOLDERS
We’ve launched better ways for you to create together - folders can now be shared! Invite people by username as Collaborators or Viewers. Collaborators can generate right into the folder and build on each other's work. Viewers can browse the images within the folder. You can also share a folder by link. Keep it private to the people you add, let anyone with the link view it, or open it up so any Midjourney member can contribute. The owner can delete a shared folder for everyone. The images stay in each creator's own history.
One heads-up: sharing doesn't change image privacy. Non-stealth images can still appear on Explore and on profiles, even if you made them inside a private folder. This is also an early version, and we'd love to hear your feedback.
WHAT YOU SAID
No more edits stuck on 'Submitting...': stalled submits now fail cleanly within 2 minutes, and jobs that secretly succeeded show up in your feed. The inline editor no longer crashes to the error screen after an edit. Keyboard shortcuts no longer fire underneath an open dialog or zoomed image. Older Firefox versions work again. And catching up on a very long history no longer stalls partway. Clicking the '+N more' pill on an image applies those settings as intended. And Live Style Previews now work on custom styles.
ORGANIZATION IMPROVEMENTS
You can pin a folder to the top of your sidebar. Search and Filters on Organize have been moved into the feed's top row where you'd expect them. Deleting a personalization profile gives you an Undo instead of an are-you-sure dialog. The pills you set in a folder stay with it, even across sessions. You can rename a folder in-place easily.
MOODBOARDS
Large uploads now happen a few images at a time with a clear progress bar, so adding multiple references no longer stalls.
PROMPT BAR & VIDEO
Using --tile with an attached image now tells you they're incompatible. Extend and Loop on video inform you the length cap if you hit it.
POLISH & BUGS
Draft mode on V8.1 and V8.2 always makes a 24-image grid, and now the tooltips, hover text, and Try buttons make that clear. The help, billing support, and submit-a-ticket links now take you where they're supposed to. Rare job types show up normally instead of crashing the page. Grid spacing got a cleanup pass. If something's wrong with your account, you now find out plainly. Temporary suspensions come with an Appeal button, and restricted accounts see dimmed controls with an explanation instead of a moderator error.
UP NEXT
Edit history that sticks around when you leave and come back. More ways to explore, find, and create styles.
Please keep sharing your feedback with us so we can continue to improve! And, as always, thank you for being part of our community ❤️
Love,
Original source
Midjourney - Oct 6, 2026
- Date parsed from source:Oct 6, 2026
- First seen by Releasebot:Oct 14, 2026
Nano Banana 2.1 on Runway Dev
Runway API adds Nano Banana 2.1 image generation with reference images, multiple output sizes and aspect ratios, and adjustable thinking levels.
October 6th, 2026 - Nano Banana 2.1 (nano_banana_2.1) is now available on Runway Dev. Generate images from a prompt of up to 5,500 characters, with up to 14 reference images (optional prompt-referenceable tags) and outputCount from 1–4. ratio is a width:height size across 1K, 2K, and 4K — 42 sizes covering 14 aspect ratios (1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, 1:4, 4:1, 1:8, 8:1). thinkingLevel is minimal, medium (the default), or high. Billed at 4 credits per 1K image, 6 per 2K, and 12 per 4K, multiplied by outputCount. Use the text to image endpoint to get started.
Original source - Oct 5, 2026
- Date parsed from source:Oct 5, 2026
- First seen by Releasebot:Oct 6, 2026
Gift Any Tier: They Pick What It Does
Civitai adds flexible membership gifting, letting users send any tier to anyone with no tier matching required. Gifts wait on the recipient’s membership page to be accepted, with clear handling for same, higher, lower, or no membership cases.
You can gift any membership tier to anyone now. No more matching the tier they already have.
- They choose when it starts. A gift waits on their membership page until they hit Accept, and it tells them what it'll do first.
- Same tier as theirs: their next months are free.
- Higher tier: they move up, free for every month you paid for. After that it renews at the new tier unless they change it.
- Lower tier: its value comes off each of their next bills. Nobody gets moved down.
- No membership: free months that end on their own, no charge.
- Cancelling partway doesn't waste it. Unused months carry on as a free membership.
Gifts already applied stay as they are.
Got a gift waiting? Check your membership page.
Original source - Oct 4, 2026
- Date parsed from source:Oct 4, 2026
- First seen by Releasebot:Oct 5, 2026
Challenge Prizes: Claim Them and Pick Your Buzz
Civitai adds a new prize claiming flow for Challenge and Crucible winners, with win notifications, a prizes page for unclaimed rewards, and automatic green payout after 30 days if you do nothing.
Challenge winners now claim their prize instead of having it dropped straight into their account. Crucible winners work the same way.
- You get a notification when you win. It says "Claim your prize!" and takes you straight to it. Everything you haven't claimed yet is on your prizes page.
- On civitai.red you pick the Buzz. Green or yellow, your call.
- On civitai.com it pays green.
- Nothing is lost if you forget. A prize you don't claim pays out in green after 30 days.
- Entry prizes haven't changed. Challenge entry prizes still pay Blue Buzz automatically.
Won something recently? Check your notifications.
Original source - October 2026
- No date parsed from source.
- First seen by Releasebot:Oct 4, 2026
Ideogram 4.5: The most precise edit model.
Ideogram adds Ideogram 4.5, reducing edit drift for cleaner multi-turn image edits. It preserves details, handles precise color and restoration changes, and supports high-resolution edits without downsizing for sharper results and seamless crops.
With each edit, image models can introduce pixel shifts, color changes, and texture artifacts. Ideogram 4.5 reduces this drift, preserving details across multi-turn edits.
MODEL COMPARISON
Compare models across multi-turn edits.
Here's a side-by-side comparison of the same edits on Ideogram 4.5 vs. GPT Image 2.5 Sunburst, Nano Banana Pro, and Nano Banana 2. GPT Image and Nano Banana outputs become unusable within a few edits, while Ideogram 4.5 stays clean edit after edit.
Precision for every kind of edit
Explore the kinds of edits Ideogram 4.5 handles with exceptional precision, from refining color and restoring old photos to making targeted changes while preserving the rest of your image.
Edit at any resolution
Edit high-resolution images without downsizing them. Ideogram 4.5 preserves the edges of an edited crop, so you can stitch it seamlessly back into the original. Create new colorways or fix a detail while keeping the rest of your image sharp and untouched—ready for close-ups and large-format print.
Source image: 4,016 x 6,016 pixels (24.2 MP)
Make your next image your best yet.
Start with an image you love. Tell Ideogram 4.5 what to change. See how far your next idea can take it.
Original source