[Definitive Guide] Summary of AI Video/Image Contests with October-November Deadlines

Introduction: The deadlines are gridlocked

I thought autumn was supposed to be more about reading, sweet potatoes, and that sort of thing.

In reality, it’s all about deadlines.

In October and November, AI video and image contests are truly gridlocked. When I tried writing them on my calendar, the text in the October 31st box overlapped so much I couldn’t read it (lol).

So, here is my current situation.

First, the AZ8 x BytePlus ‘Time Machine Launch Plan’. On October 12th, I posted a commercial for a fictional product called ’10-Second Overtime’ on X. It’s about ‘Modrop,’ an edible time machine that lets you go back 10 seconds with one bite. It’s a story about what if you could redo a penalty kick for just 10 seconds… By the way, the deadline for registration and credit submission for this contest was October 7th. Since I’ve already registered, I should just need to finish the work and submit it… I’ll double-check the ‘post to the AZ8 community’ requirement before the deadline.

By the way, I also entered AZ8’s previous ‘Ad from Zero’ contest with ‘-1g ROPE’ and ‘STORY RUNNER,’ and the results were announced on October 11th. …My stomach hurts just writing this myself.

Second, JACC’s new MV contest ‘AI MUSIC VIDEO WAVE 2026’. Starting October 13th, with a total prize pool of 400,000 yen and 13 categories. As someone doing 100 MV knocks, I can’t afford not to enter. The application requirements were finally released on the night of October 13th. The deadline is November 3rd (Tue, Holiday) at 23:59, the length must be within 180 seconds, and you can use any tools. I’ve been reloading the page since morning so much that the print on my F5 key is starting to fade, so I’m honestly relieved (lol). I even tweeted on X, ‘Do I have to use Topview?’, but the requirements say ‘Any tools allowed.’ There is a TopviewAI Award (50,000 yen), but there are no tool restrictions just for entering.

Speaking of JACC, that’s where I won the semi-grand prize for my ‘Unreasonable Generator for Gemini Omni.’ The contest where I made a monk do 108 bench presses. Ever since then, I’ve felt a personal debt of gratitude to JACC (lol).

So, while I was at it, I researched every contest with an October-November deadline. And this time, just researching wasn’t enough, so I added 3 project ideas for ‘how I would submit’ to every single contest. That’s 30 ideas in total. …Complaining at the beginning that the deadlines are gridlocked while adding 30 more projects for myself is pretty crazy, right? (lol) But honestly, thinking about ‘this fits in this length’ or ‘this judge will look for this’ while reading the rules is my favorite time.

Feel free to steal these project ideas. Actually, if someone submits using the same idea, I’d love to see it. I’ve included all the source URLs. Bookmark this and look back at it the night before the deadline.

Comparison Table

Difficulty is rated 1-5, based on my subjective opinion. It’s based on length, strictness of rules, tool restrictions, and the intensity of the competition.

One by one, honestly (+ how I would submit)

First, how to read the project ideas. Proposed Title / One-line Synopsis / Potential AI Tools / Points that might catch a judge’s eye. I’ve written them in these four categories. The tool names mean ‘this is what I would try,’ so please check each tool’s terms of service before using them to ensure they meet the requirements (commercial use, output size, vertical support, etc.).

AICU Award 2026 (Phase 2 Deadline: 10/16 21:00)

This is the closest deadline. It’s this week. Since it’s an international award, I think this is the place to polish and submit ‘the strongest piece you have right now.’ It’s open to works published between January 1, 2025, and December 31, 2026, that can be viewed without logging in. In other words, you can submit works that have already been published. It doesn’t matter if you’ve entered or won other contests before. They are currently accepting entries for Phase 2, and the deadline is October 15th AoE, which is October 16th at 9:00 PM Japan time. As of 11:00 PM on October 12th, they had received 169 entries from 99 people across 12 countries and regions. There are 10 categories, and after a qualification review, there will be a public vote (October 26th – November 15th), followed by a final selection by judges with critiques. The awards ceremony is on November 20th at the INTER BEE IGNITION x DCEXPO stage at Makuhari Messe. I couldn’t find any mention of prize money on the official page, and winners receive a certificate that can be verified. It’s a place to go for a ‘title’ rather than money.

And, upon re-reading this time, I noticed something I missed. There is a submission fee of USD 10 per entry (free if you have a partner or invitation code from a past contest). I didn’t write this in my previous draft, so I’m correcting it here. I’m sorry. There’s no limit to the number of entries, but if you submit 3, that’s 30 dollars. For my wallet, which keeps burning through credits, that’s a decent amount (lol).

How I would submit (ideas on ‘which category to place’ existing works)

Idea 1: Submit ‘Strike the Road’ to the ‘Music, MV, Dance’ category – One-line synopsis: A Japanese-style dark fantasy MV. I’ll put my current signature work on the shelf of an international award just as it is. – Potential AI tools: The main video is already produced. The required representative images (1:1, 16:9 cover, 16:9 without text) can be cut from the main video or recreated using image generation like Nano Banana 2.1. – Points that might catch a judge’s eye: This category requires inputting music/performance credits and lyrics, so the minimum requirement is to have the music rights explained (this is per the official rules). The final judges include Yohei Sadojima (Cork) and Matty Shimura (ElevenLabs’ Chroma Awards), and the method involves a critique for every vote. So, I think the key is whether I can clearly explain ‘why it’s Japanese-style’ in a 500-character summary (this is my guess).

Proposal ②: ’10-Second Overtime’ for the ‘Advertising/PV/Broadcast’ category (spec work) – One-line synopsis: A fictional commercial for ‘Modrop,’ a candy that lets you go back 10 seconds. A story about a man redoing a penalty kick for just 10 seconds. – AI tools to use: Already produced in AZ8 (per AZ8 regulations, video generation must be within AZ8). The representative image is made from cuts exported from the AZ8 production canvas. – Points likely to hit in judging: This category requires a client name, and the official rules state that for independent work, you can just write ‘spec work.’ A commercial for a fictional product proves that ‘you can do advertising planning with AI,’ so it should be easy for judges in the advertising category to understand (speculation). However, since the AZ8 results are scheduled for 10/26, check the AZ8 terms to ensure it doesn’t count as a double submission.

Proposal ③: ‘Unreasonable Request Generator for Gemini Omni’ for the ‘Open’ category – One-line synopsis: A JACC semi-grand prize winning work where a monk is made to do 108 bench presses. – AI tools to use: Gemini Omni (already produced). – Points likely to hit in judging: The Open category requires a field to write ‘what is new’ (per official regulations). If you can assert that ‘the way of playing with unreasonable requests itself is new,’ I feel it will fit into the ‘category that doesn’t fit anywhere’ (speculation). Before submitting, check the official regulations to see if the public URL can be opened without logging in and if the publication date is within the target period.

Honestly, if I submitted all three, it would cost 30 dollars. I’ll stick to just ① and submit the rest in my heart (lol).

Source: https://aicu.ai/award/2026/ / https://aicu.ai/award/regulations/index.ja.html (Regulations, mandatory items by category, judging flow) / https://note.com/aicu/n/n049f63ab41ec (2nd deadline, entry fee USD 10, final judges, schedule. Published 2026/10/12)

Recommendation level: ★★★ (Only for those who already have works. Given the entry fee, narrowing it down to one is also an option)

AZ8 x BytePlus ‘Time Machine Launch Plan’ (Deadline 10/21)

This is the one I’m submitting. The 2,000 dollar gold prize is huge. However, the regulations are quite heavy. 60 seconds or longer, Japanese and English subtitles are mandatory, all AI video generation must be within AZ8, and the production canvas must be made public. A designated watermark is also required. You can use any AZ8 model, but only BytePlus models can use free points. Judging is 35% advertising expression, 25% creative concept, 25% visual, and 15% story. If there’s a tie, it’s decided by the advertising expression and concept points. In other words, making people think ‘I want a time machine!’ is stronger than ‘the video is beautiful.’ And importantly: Registration and credit applications closed on October 7 at 23:59. The official flow is ‘Application form -> Production -> X post -> AZ8 community post,’ so check the official page to see if those who haven’t applied can still submit now. These work proposals are for those who have already applied and for my future self.

If I were submitting (for those who have already applied)

Proposal ①: ’10-Second Overtime — Complete Version’ – One-line synopsis: A man who missed a penalty kick goes back 10 seconds with ‘Modrop’ and kicks again. But in the end, regret that can’t be undone in 10 seconds remains. – AI tools to use: Video generation model within AZ8 (AZ8 exclusive per regulations). Subtitles are in two lines, Japanese and English, added with external editing software. *Check the official application requirements to see if it’s okay to add subtitles with external editing software. – Points likely to hit in judging: The 35% weighted advertising expression is ‘Does it make you want to own it?’ (per official judging criteria). So, I want to end the last 5 seconds not with the story, but with the ‘face of the product,’ like ‘◯ pieces per bag.’ The current version leans too much on the story, so I’m debating whether to fix this myself.

Proposal ②: ‘Yesterday Pass — A Subscription Time Machine to Go Back Just 24 Hours’ – One-line synopsis: A subscription-type time machine that lets you go back to ‘yesterday’ once a month. The use case could be redoing a confession or taking back a word you said too much. – AI tools to use: Video generation model within AZ8 + image generation to create an app-style UI cut (it’s safer to do image generation within AZ8 as well). – Points likely to hit in judging: The 25% concept is ‘novelty and memorable elements’ (per official judging criteria). Betraying the ‘time machine = giant machine’ trope and making it a smartphone subscription. Showing a price list should also affect the 35% advertising expression.

Proposal ③: ‘One-Way Ticket — To Grandma’s Kitchen’ – One-line synopsis: A ticket that lets you go to the kitchen where your late grandmother is cooking just once. Return tickets are not sold. – AI tools to use: Video generation model within AZ8 (to keep the same character, fix the reference image and rotate). – Points likely to hit in judging: The 15% story is ‘is the motivation clear and are the emotions natural?’ (per official judging criteria). The weighting is small, but a proposal that aims to make people cry is unlikely to overlap with others. However, if it leans more toward ‘tear-jerker’ than ‘I want it,’ it will drop the 35% advertising expression, so whether you can show the ‘price of the ticket’ at the end is the deciding factor.

…and as I wrote this, I realized. If I remake three works from now, I won’t have time to add subtitles on the 21st (lol).

Recommendation level: ★★★★ (Those who have applied, give it your all. Those who haven’t, prepare for next time)

JACC ‘AI MUSIC VIDEO WAVE 2026’ (Deadline 11/3 23:59)

Music x Video x AI. Total prize money of 400,000 yen, with 13 categories to be awarded. Having many categories means it’s not a winner-takes-all grand prize, but 13 chances to ‘hit it big with something.’ The key visual catchphrase is ‘Let’s create the next wave.’ The main sponsor is Topview AI (listed as a Gold Sponsor on the official site). The first round of judges was announced on 10/12, including former Avex director Motohiro Abe, former PENICILLIN member GISHO, Reiji from THE SIXTH LIE, Yuyu Nagao, and WIZAI representative Koichiro Tanaka. …Wait, real people from the music industry are going to watch my MV? I’m trembling (lol). By the way, to the question from overseas people ‘Can non-Japanese people submit?’, the official replied ‘Of course, OK, we will issue announcements in both Japanese and English.’

The application requirements were released in an official X post at 19:00 on 10/13. Here are the key points: – Period: 10/13 (Tue) 19:00 – 11/3 (Tue/Holiday) 23:59. Results announcement scheduled for 11/19 (Thu) – Length: Within 180 seconds. However, accounts without X Premium can only post up to 140 seconds, so those going for long-form should check their own account first – Music: Self-made or AI-generated (commercial use OK). Existing songs, covers, and sampling are NG, and AI-mimicking the voices of real singers or voice actors is also NG – Eligibility: The official X announcement says ‘No restrictions on age, nationality, or residence (minors need parental consent),’ but the official website says 18 or older, so be sure to check the official source – Application: ① Like the post + follow @JACC_AI ② Quote repost the post with the work video and #AIMUSICVIDEOWAVE2026 ③ Bookmark the post. Once JACC ‘likes’ it, the submission is complete – No limit on the number of entries. You can submit as many as you want (this is a dangerous sentence for the ‘100 MV knock’ crowd) – Judging points are four: ‘Video Expression,’ ‘Music,’ ‘Unity of Music and Video,’ and ‘Narrative.’ Judges also include Beat Flickers, MINIMA, AI FREAK, and Mankyu.

The 13 categories have also been released. In addition to the Grand Prix, Topview AI Award, Euclid Solutions Award, and Jury Special Award, there are 9 genre-specific ones. Most Supported Award, Video God Technique Award, Floor Frenzy Award, Enchanting Heroine Award, Emotional Award, Visual Kei Gothic MV Award, Shocking MV Award, Artistic Award, God OP/God PV Award. …Did you make the ‘Shocking MV Award’ for me? (It’s not)

If I were submitting

Proposal ①: ‘Strike the Road, Second Strike’ – One-line synopsis: A sequel to ‘Strike the Road.’ The sound of the blacksmith striking the sword becomes the beat, and every time he strikes, the road extends one by one. A Japanese-style dark fantasy. – AI tools to use: Vidu Q4 (fix character and costume with 15 reference images, 16 seconds max per take, so calculate about 9 takes for 140 seconds), MiniMax H3, and Kandinsky 6.0 Video only for the final lip-sync (5 seconds with audio). Tools are free, so if you want to aim for the Topview AI Award, you could mix in Topview. The song is newly created for the sequel (self-made or AI-generated, commercial use OK). If you reuse the song from the previous work ‘Strike the Road,’ check the official application requirements to see if it meets the music conditions. – Length: Within 180 seconds. If my account isn’t X Premium, I’ll keep it within 140 seconds. – Points likely to hit in judging (per official judging points, category selection is my assessment): ‘Unity of Music and Video’ is clearly included in the judging points. Matching the blacksmith’s strikes to the beat is exactly what that aims for. For the category, God OP/God PV Award or Video God Technique Award.

Proposal ②: ‘Song of 17 Views’ – One-line synopsis: Songs that were hardly seen by anyone in the 100 MV knock gather in the middle of the night to sing in a chorus. – AI tools to use: Characterization with MiniMax H3 or Vidu Q4, showing lyrics with JIZURA (a text PV tool usable in a browser, MIT license). The choral song is self-made or AI-generated, using one that is commercial use OK. The singing voices of the ‘rejected songs’ also do not mimic real singers or voice actors. – Length: Keep within 140-180 seconds (up to 140 seconds if no X Premium). – Points likely to hit in judging (per official judging points, category selection is my assessment): ‘Narrative’ is a judging point. Aim for the Emotional Award or Shocking MV Award with the temperature gap of closing self-deprecating jokes with a serious chorus. *’17’ is a performance. Replace with the actual number.

Idea 3: ‘The Impossible Request MV: The Monk, 108th Chorus’ – One-line synopsis: Reincarnating an ‘impossible request’ that won a runner-up prize into an MV. The monk sings the chorus to the rhythm of 108 bench presses. – Potential AI tools: Gemini Omni (for the impossible request), video generation using Vidu Q4 with reference audio (up to 3 clips) to maintain the same voice. The monk’s singing voice will be an original voice created by AI, not imitating any real singer or voice actor (which is prohibited by the rules). The monk will also be a fictional character (real people’s faces or names cannot be used without permission). The song will be original or AI-generated and cleared for commercial use. – Duration: Doing all 108 reps would definitely exceed 180 seconds, so I’ll structure it to reach the 108th rep during the chorus to keep it under 140 seconds (lol). – Points likely to hit with judges (beyond official categories): There really was a ‘Shocking MV Award.’ I’m going for this one seriously.

The deadline is November 3rd. I will narrow it down to one of these three ideas and write about the production process on note. I’ll also hold a memorial service for the two rejected ideas (lol).

Sources: https://x.com/JACC_AI/status/2109949689500213663 (X post for application requirements) / https://x.com/JACC_AI/status/2109590334321598887 (Judge announcement/key visual) / https://x.com/JACC_AI/status/2109513578676179218 (Reply confirming overseas participation is OK) / https://jacio.or.jp/terms (Application terms) / https://jacio.or.jp/aimvwave2026/ (Official recruitment page)

Recommendation Level: ★★★★★ (All MV creators, assemble. I’m entering too.)

Re-ROBO AI SHORT MOVIE FESS 2026 (Deadline: 10/31 12:00)

Grand prize is 300,000 yen. Plus, the top 24 works will be screened on the big screen at the Akihabara UDX Theater (November 14th, free admission). My own work playing at cinema size. Just imagining it makes my palms sweat. The theme is ‘The Future 14 Years From Now.’ You choose a sub-theme from ‘Social Issue Resolution’ or ‘Free Style,’ and submit in one of four categories based on age group (17 or under / 18 or over) x genre (Robot/Drone / SF/Fantasy). Under 5 minutes, up to 2 entries per person, free to enter. The condition is that AI must be used in a way that ‘substantially contributes to the work’s expression,’ and you must list the tools and processes used. Narration and combinations with live-action footage are OK. Judging is based on creativity, technical skill, and expressive power. The winner will be decided on the day by combining judge scores and audience popularity votes. The head judge is Keiju Kure, representative of Creators Match (other judges to be announced later). Note that the deadline is 12:00 PM. It’s not at night. I’m the type who would definitely misunderstand this somewhere, so I’m writing it here in bold.

One thing that caught my attention is that both genre category names are ‘Anime Category.’ Check the official application requirements for details on whether you can submit live-action-style AI footage.

If I were submitting

Idea 1: ‘2040, The Last Delivery Drone’ (Robot/Drone / Social Issue Resolution) – One-line synopsis: A delivery drone goes to deliver the final package to a mountain village where people are disappearing. The recipient is a grandmother who no longer has anyone to write letters to. – Potential AI tools: Vidu Q4 (fixing the drone shape with reference images), MiniMax H3, narration can be my own voice. – Points likely to hit with judges: Since ‘Expressive Power = Expression of the Theme’ is a judging criterion, I will clearly show 2026 + 14 years = 2040 on screen (from official judging criteria). Depopulation x logistics is easy to explain as a social issue.

Idea 2: ‘To Me, 14 Years From Now, From AI’ (SF/Fantasy / Free Style) – One-line synopsis: The AI I created in 2026 replies to me 14 years later. The content is a lecture that’s a bit hard to hear. – Potential AI tools: ChatGPT (GPT-6) for script generation, MiniMax H3 for video, Kandinsky 6.0 Video for just the one-cut lip-sync. – Points likely to hit with judges: The ‘Creativity’ judging criterion includes ‘ideas that leverage AI’ (from official judging criteria). A structure where the AI becomes the narrator itself is an idea that hits that head-on.

Idea 3: ‘Akihabara 2040 Karakuri Festival’ (SF/Fantasy / Free Style) – One-line synopsis: Akihabara in 2040. In the alleys of the electric town, Edo-period mechanical dolls and AI robots hold a festival for one night only. – Potential AI tools: Vidu Q4 (unifying the mechanical design with 15 reference images), Seedance (fixing characters with three-view drawings). – Points likely to hit with judges (speculation): In the end, audience popularity votes at the Akihabara venue will count toward the score. A work set in the very city where it’s being screened should elicit ‘Oh, it’s here!’ reactions at the venue. As someone who likes Japanese-style themes, this is the idea I most want to submit.

Recommendation Level: ★★★★ (Serious about both the prize money and the stage)

VISIONARIA (Rome / Regular 10/31)

It’s an overseas film festival, and the fact that there is an AI Music Video category is a big plus for me. There are 6 categories (Film, Animation, Documentary, Music Video, Experimental, Short Form). There is a €10 submission fee via FilmFreeway. Plus, there is no premiere requirement. You can submit works that have already been uploaded to the internet. And the prize money. The terms state that ‘unless otherwise specified, prizes do not include cash,’ and as of now, no cash prizes have been announced. In other words, you pay €10 to go for the honor of potentially being screened in Rome (lol). English subtitles are required for dialogue in languages other than English or Italian. As someone doing Japanese MVs, I plan to get through it with translation tools and sheer willpower. Selection is based on the film festival’s selection committee and judges’ decisions, and no scoring criteria like a scorecard are public. Therefore, the ‘points likely to hit’ below are all speculation. I couldn’t find a maximum duration either, so check the official application requirements for details.

If I were submitting

Idea 1: ‘Strike the Path (English subtitled)’ -> Best AI Music Video – One-line synopsis: Add English subtitles to a previously released Japanese-style dark fantasy MV and head straight to Rome. – Potential AI tools: The main feature is already produced. I’ll have ChatGPT do a rough translation of the English, then check and correct the meaning of the lyrics myself. – Points likely to hit with judges (speculation): The MV category description is ‘MVs that used AI to create concepts, visuals, animation, and audiovisual expression.’ ‘Japanese style’ at an Italian venue should stand out quite a bit just by itself.

Idea 2: ‘100 MV Knockout’ -> Best AI Documentary – One-line synopsis: A record of a man trying to make 100 MVs with AI, only to be buried in a mountain of melted credits and rejected cuts. – Potential AI tools: Recordings of my own production screen + rejected cuts, Claude for organizing the structure, ChatGPT for rough English subtitles. – Points likely to hit with judges (speculation): Since this is a film festival where you can’t submit unless AI is the ‘center of the work creation,’ making the process of struggling with AI the subject itself makes sense. Honestly, I don’t know if a self-deprecating documentary will work overseas until I try submitting it (lol).

Idea 3: ‘Ukiyo 15 Seconds’ -> Best AI Short Form Content – One-line synopsis: A vertical short where Ukiyo-e paintings come to life for just 15 seconds, and the waves, Mt. Fuji, and people all return to paper in the final second. – Potential AI tools: Vidu Q4 (since it’s max 16 seconds at a time, it’s almost a one-shot export), image generation for creating Ukiyo-e style original art. *Need to check rights regarding whether to use existing Ukiyo-e works as input directly. – Points likely to hit with judges (speculation): Check the official site for the duration and aspect ratio conditions for the Short Form category. I believe a structure that ‘has a beginning and an end within a short time’ is easy to convey to judges in any country.

Sources: https://filmfreeway.com/Visionaria-festival (Schedule, categories, terms) / https://aifilmcontests.com/contests/visionaria-rome-2026 / https://aifilmcontests.com/topics/ai-film-festivals-in-italy-2026 (No premiere requirement, subtitle condition summary)

Recommendation Level: ★★ (For those who want to try submitting overseas once)

Sabae eSports DAY AI Illustration Contest (Deadline: 11/6)

It’s an illustration contest, not video. You can submit via X, up to 3 entries per person. The theme is “A New Worldview of Sabae City and eSports,” and the winning entry will be used as the main visual for the “Sabae eSports DAY” on December 5th. Results will be announced on November 20th. The Grand Prize is a pair of eSports glasses made in Sabae, the “City of Glasses.” The local flavor of the prize is fantastic (lol). The format is 4:5 vertical, with a recommended size of at least 2,894px wide by 3,617px high. Conditions include using commercial-use-permitted generative AI tools according to their terms, not imitating specific existing works or characters, and saving your prompts and reference images. Do not delete your generation logs. The judging criteria have not been made public, so the “points that might hit the mark” below are speculation.

By the way, Sabae isn’t just about glasses. While it accounts for over 90% of domestic eyeglass frame production, the Kawada district is also a production center for Echizen lacquerware, which holds over 80% of the national market share for commercial lacquerware. This was the part that got me most excited while researching.

If I were submitting

Idea 1: “The Finals Inside the Lens” – One-line synopsis: The reflection of the final match arena and the Sabae mountain range is visible in the lens of the glasses worn by a boy in the spectator seats. – Potential AI tools: Create the composition with GPT-Image-2.5 or Nano Banana 2.1, then upscale to the recommended size at the end. Check the terms of each tool for commercial use conditions. – Points that might hit the mark in judging (speculation): Since it’s a call for the event’s main visual, it’s important that “glasses” and “eSports” are immediately recognizable in a single image. Making the spectator, rather than the player, the protagonist makes it easier for attendees to project themselves into the scene.

Idea 2: “The Jet-Black Controller” – One-line synopsis: An Echizen lacquerware artisan is finishing a game controller decorated with maki-e. On the monitor in the background, a tournament is about to begin using that controller. – Potential AI tools: GPT-Image-2.5 (for drawing fine patterns), with manual color adjustment for the finish. – Points that might hit the mark in judging (speculation): Expressing the “new worldview” not just through glasses, but through lacquerware—another face of Sabae—makes it less likely to overlap with other entries. Everyone is definitely going to do the glasses angle (lol).

Idea 3: “Glasses Bridge Stadium” – One-line synopsis: A future Sabae where the frames of glasses have become a massive bridge, and the lenses on both sides have become stadiums. – Potential AI tools: Nano Banana 2.1 for the bird’s-eye view of the city, and GPT-Image-2.5 to add and composite the people. – Points that might hit the mark in judging (speculation): Looking at the gallery of submitted works, ideas involving glasses x arenas are already starting to appear. Therefore, I feel that fleshing it out into a “worldview that functions as a map” will have a stronger appeal when displayed.

Sources: https://aicompe.com/contests/sabae-ai-illustration-contest-2026 (Application Guidelines) / https://www.echizen.or.jp/cooperative (Echizen Lacquerware Cooperative: Kawada district, over 80% share of commercial lacquerware) / https://renew-fukui.com/about/monozukuri/ (Over 90% of domestic eyeglass frame production)

Recommended level: ★★★ (A breather between videos. But I really want those glasses.)

AI Video Submission Festival 2026 (11/1 – 11/15)

A project with a very wide entry point: “As long as you have an AI-generated thumbnail, the main content doesn’t have to be AI-made.” The rules are: register the announcement video (sm46579758) as the parent work, lock the tag “AI Video Submission Festival 2026,” use an AI-generated thumbnail, show the “names of the AI used” and the “thumbnail prompt” within the video, and include a link to the YouTube version in the description. Up to 3 entries per person (same video on Niconico and YouTube). Judging is handled by Google’s Gemini, and it’s said to be a “pure meritocracy” that ignores view counts and likes entirely. Scores and evaluation comments will be published for all works when results are announced. It seems a prompt for judging is also being distributed, but I haven’t opened the distribution file this time. Therefore, the “points that might hit the mark” below are speculation. For those participating, I think it’s safest to read the distributed prompt yourself before creating your entry. The prizes are modest, so I recommend this as a practice match for those entering a contest for the first time.

If I were submitting

Idea 1: “I had an AI judge grade my rejected MV” – One-line synopsis: A live commentary where I have Gemini grade an MV I rejected during my “100-knock” challenge, and then I fix it exactly as told. – Potential AI tools: Gemini (for self-checking beforehand), thumbnail by Nano Banana 2.1. – Points that might hit the mark in judging (speculation): Since Gemini is also the judge, a structure that shows “how I accepted the AI’s feedback and made corrections” should make it easy to write evaluation comments. However, check with the officials just in case to ensure pre-checking isn’t a rule violation.

Idea 2: “I’ll show you all the melted credits” – One-line synopsis: An accounting report video that counts how many times I generated, how many credits I melted, and where I cried while making one MV. – Potential AI tools: The main content can be screen recording without AI (the rules allow for non-AI main content). Only the thumbnail is “a picture of stacks of cash melting” made with GPT-Image-2.5. – Points that might hit the mark in judging (speculation): If the judging ignores numbers entirely, even someone like me with zero views is on the same playing field (lol). I think videos that compete on the density of information are compatible with AI judging.

Idea 3: “A 1-minute hand-drawn animation with an overly serious thumbnail” – One-line synopsis: I make only the thumbnail look like a masterpiece using AI, while the main content is my own poor hand-drawn flipbook animation. But in the last second, it catches up to the thumbnail’s art. – Potential AI tools: Thumbnail by Nano Banana 2.1, main content hand-drawn. – Points that might hit the mark in judging (speculation): An idea that turns the “AI thumbnail required, main content free” rule—unique to this project—on its head. The Q&A states that “screenshots of chat AI responses may result in point deductions,” so I’ll avoid that kind of laziness.

Source: https://www.nicovideo.jp/watch/sm46579758 (Organizer’s announcement video and description)

Recommended level: ★★ (For first-time participants)

SKY PerfecTV! 30th Anniversary Short Video Contest (Deadline: 11/15)

The Grand Prize in the Creator Category is worth 100,000 yen in Amazon gift cards, and the 30th Anniversary Special Prize is worth 50,000 yen. It must be a vertical video of at least 15 seconds, and you must include “SKY PerfecTV!” somewhere in the video. Submit via TikTok or Instagram, and the Creator Category also requires filling out a dedicated form. The theme is “Capture what you love.” Genres and expressions are free; movies, anime, documentaries, or experimental videos are all fine. The judging criteria are clearly published: “Thematic expression,” “Empathy/ability to move emotions,” “Immersion (whether you want to keep watching after the first few seconds),” and “Uniqueness,” with “Video quality” added for the Creator Category. The judges are “currently being selected.” Now, about AI. I read all the official application guidelines and terms, and it doesn’t say “OK” or “NG” regarding generative AI. What is required is that it be an “original work” and “not infringe on the rights of third parties.” It’s scary to assume that “not written = OK,” so if you’re submitting with AI, it’s safer to check with the secretariat first. The project ideas below are also written on the assumption that AI is confirmed to be OK.

If I were submitting (once AI usage is confirmed to be OK)

Proposal 1: ‘2:00 AM, I Screamed Alone’ – One-line synopsis: A person in a pitch-black room, screaming at the goal in front of the TV alone. That voice overlaps for just an instant with tens of thousands of people on the other side of the screen. – AI tools to use: MiniMax H3 or Vidu Q4 for room shots, exported in 9:16. Do not use real athletes or match footage (it will trigger rights regulations). – Points likely to hit in judging: A proposal aimed at the ’empathy/ability to move emotions’ judging criteria (based on official criteria). Since the official message also mentions ‘the night I screamed at my favorite athlete’s goal,’ I shouldn’t misread the theme.

Proposal 2: ‘Resolution of Love’ – One-line synopsis: It starts with blurry image quality, and as the ‘love’ deepens, the resolution increases, until it is clearly visible in 4K at the end. – AI tools to use: Vidu Q4 (decide the composition with low resolution, and output only the final cut in high resolution). – Points likely to hit in judging: ‘Whether you want to keep watching in the first few seconds’ is included in the judging criteria (based on official criteria). If the beginning is blurry, people will stop scrolling thinking, ‘What is this?’ This is a vertical-oriented proposal that can show the difference between the beginning and the end in 15 seconds.

Proposal 3: ’30 Years of Remote Controls’ – One-line synopsis: In a family living room, 30 years’ worth of remote controls are replaced with new ones. But the voices fighting over the channel remain the same. – AI tools to use: Vidu Q4 (fix the family’s faces with reference images and change only the era), and a ‘SKY PerfecTV!’ caption at the end. – Points likely to hit in judging (speculation): Since there is a ’30th Anniversary Special Award’ in the creator category, a proposal that carries 30 years might reach it. However, it is not written what they are looking for to select the award, so this is speculation.

Source: https://www.skyperfectv.co.jp/30thanniversary/shortmovie/guideline/ (Theme, Judging Criteria, Prizes) / https://www.skyperfectv.co.jp/30thanniversary/shortmovie/terms/ (Terms)

Recommendation Level: ★★ (★★★★ if AI eligibility is confirmed)

AICU ‘#Synthetic Reality Moratorium MessageVideo’ (Deadline 11/16)

A project to create an MV for the designated song ‘Synthetic Reality Moratorium’ (Japanese version) or ‘Moratorium in Synthetic Real’ (English version). It doesn’t have to be the full length; 60 seconds or more including the chorus is fine. Post it on YouTube or X with #SyntheticRealityMoratorium and apply from the AICU page. Results will be announced on November 30. The total prize is 500,000 AP (Grand Prize 200,000 AP, Excellence Award 100,000 AP x 2, Selected 20,000 AP x 5). Note that these are points usable at AICU, not cash. There is also an ‘AICU API Award’ selected from works using AICU API or AICU Create. The theme is ‘I don’t want to use AI. I understand that feeling, too.’ The recruitment page clearly states that they are looking for an MV that makes people who think ‘I don’t want to use AI’ feel ‘If this kind of expression is possible, maybe it’s okay.’ There is no grading sheet, but I read this as the actual axis of the judging. Song arrangement is not allowed (excerpts are okay). Sometimes copyright claims appear on YouTube; if they do, consult AICU before filing an objection. How do we, who create with AI, respond to this question? Honestly, this is the project that makes me think the most.

If I were submitting

Proposal 1: ‘I Won’t Erase My Clumsy Lines’ – One-line synopsis: I won’t let AI fix the shaky line art I drew by hand. The AI adds color and movement, and only the lines remain human until the end. – AI tools to use: From hand-drawn line art, use MiniMax H3 (specify the first and last images to generate the middle) or animate with Wan3.0. – Points likely to hit in judging: A proposal that responds directly to the ‘courage to use AI, and the courage not to use it, are both the creator’s own choices’ on the recruitment page (based on the purpose of the recruitment page). The fact that my drawing is clumsy becomes a weapon for the first time (lol).

Proposal 2: ‘Collection of Unused Cuts’ – One-line synopsis: A lyric video made only of cuts I generated with AI but discarded, thinking ‘this is wrong.’ At the end, the reason for discarding them overlaps with the lyrics. – AI tools to use: Discarded cuts (my own generation logs) + JIZURA for lyric text effects. – Points likely to hit in judging (speculation): Since 1 out of 5 official demos is a lyric video with only text, doing only text effects will overlap. I think showing ‘I used AI, but it was a human who chose’ makes the response to the theme one level deeper.

Proposal 3: ‘Moratorium in Synthetic Real — Two Hands’ – One-line synopsis: In the English version, a human hand and an AI hand on the left and right of the screen. They fight over a pen at first, but hold it together during the chorus. – AI tools to use: AICU Create’s MiniMax H3 (eligible for the AICU API Award), Wan3.0 for comparison. – Points likely to hit in judging (speculation): The AICU API Award is for works using AICU API/AICU Create (details of the prize will be announced separately). I feel that since few people will choose the English version, it will be easier to catch the eye just by doing that.

Source: https://aicu.jp/call/moratorium (Application Requirements, Prizes, Contest License)

Recommendation Level: ★★★★ (For those who want to win with a message)

Kyoto AI Film Festival 2026 (Deadline 11/30)

A short film of about 5 minutes or less related to Kyoto (including title and end credits, no lower limit). You don’t have to live in Kyoto; a fictional Kyoto or a future Kyoto is also fine. Previously published works can be submitted, and the screen ratio is free. Upload to YouTube as public or unlisted, and send the URL from the form on the official website. Up to 2 works per person, free entry. All submitted works will be posted on the official website, and after web voting from December 1 to 25 and a metaverse screening, the Grand Prix will be announced on ‘Cinematograph Day’ on December 28. Prize money and prizes are not written on the official website or the recruitment article. Judging criteria other than ‘web voting’ are not written either, so the ‘points likely to hit’ below are speculation. As someone who makes Japanese-style MVs, I can’t miss this.

If I were submitting

Proposal 1: ‘Hyakki Yagyō, Crossing Shijo-dori’ – One-line synopsis: 2:00 AM on Shijo-dori. A parade of yokai stopped at a red light, waiting properly for it to turn green before crossing. – AI tools to use: Vidu Q4 (unify yokai designs with 15 reference images), Seedance (fix characters with three-view drawings), and Kandinsky 6.0 Video for the finishing voices. – Points likely to hit in judging (speculation): Since it’s decided by web voting, an image that makes people realize ‘it’s Kyoto’ in the first 10 seconds is strong. Japanese-style dark is my home turf, so I’m most confident in this proposal.

Proposal 2: ‘There Is No Place for Me at Equal Intervals on the Kamo River’ – One-line synopsis: A line of couples sitting at equal intervals on the banks of the Kamo River. I, who came alone, wander for 5 minutes not knowing where to sit. – AI tools to use: MiniMax H3 (create a line on the riverbank in a wide shot), narration with my own voice full of self-deprecation. – Points likely to hit in judging (speculation): If the voters like Kyoto, they will laugh at the ‘relatable’ aspect. It’s a project that can be submitted even if you don’t live in Kyoto, so it’s a proposal that uses the outsider’s perspective as material.

Proposal 3: ‘2126, People Are Still Pulling the Yamaboko of the Gion Festival’ – One-line synopsis: Kyoto 100 years from now. In a city where everything has become automated, only the Yamaboko are still pulled by humans. – AI tools to use: Vidu Q4 (wide shot of the cityscape in 4K), crowd cuts with MiniMax H3. – Points likely to hit in judging (speculation): Since the application guidelines welcome ‘future Kyoto,’ it’s right in the middle of the theme. Create it in a way that doesn’t use real footage of the festival or logos of specific organizations (for rights reasons).

Source: https://cinepu.com/compe/gsSog3XViA/ / https://kyototext.jimdofree.com/ (Application Requirements, Schedule)

Recommended Level: ★★★★ (Japanese-style creators can easily enter with existing works)

Before you apply, please take a look at this

From here on, I’ll share things I’ve messed up or almost messed up in the past.

Check the deadline “time” as well. I thought it was 23:59, but it was 12:00. Re-ROBO was exactly this case. Some, like the AICU Award, use “AoE” (Anywhere on Earth) notation. When converted to Japan time, it might be the following night, so always convert it.

Check if there are any costs involved. The AICU Award has a submission fee of USD 10 per entry, and VISIONARIA has a submission fee of €10. Others don’t mention an entry fee. However, for JACC, you need X Premium (a paid plan) to post works longer than 140 seconds on X. If you don’t want to spend money, it’s safer to keep it under 140 seconds. This time, I missed the AICU submission fee. Read the terms to the end. I’m telling myself this (lol).

Is the duration “or more” or “or less”? AZ8 is “60 seconds or more,” Re-ROBO is “5 minutes or less,” and JACC is “180 seconds or less” (effectively 140 seconds for some due to X posting). Sometimes it’s the opposite, or there are two-stage requirements. I previously made a 1-minute PV with four 15-second clips using MiniMax H3, but if I submitted that to a “60 seconds or more” requirement and it turned out to be 59 seconds, that would be the end of it. Don’t push the limits; leave some breathing room.

Check for tool restrictions. AZ8 says “All AI video generation must be done with AZ8.” Mixing in clips made with other tools could be a violation. Conversely, AICU and Re-ROBO allow free production methods. Whether you can reuse the same work depends on this.

Rights. This is the most important part. Music, fonts, sound effects, reference images. Are they all things you are “allowed to use”? It’s written that the designated song for AICU cannot be arranged and that copyright claims might appear on YouTube. The handling of winning works (the scope of free use by the organizer) also varies greatly from contest to contest. I felt this keenly when I read the terms for the HUSTLER CM Koshien. Save your prompts, reference images, and licenses for materials used in a folder. Just doing this will keep you from panicking if you’re asked about anything.

Double submission of the same work. Some places say “works waiting for results from other contests are not allowed.” Check the terms before submitting.

Next preview: I’ll write about my challenge

As of writing this article, my October looks like this.

AZ8 is already submitted, and I’m debating whether to fix it before the 21st deadline. JACC has a 11/3 deadline. I wrote 3 ideas above, but I haven’t decided on one yet (lol).

In the next note, I’ll reveal everything from brainstorming to the final product, including how I make the JACC MV and which of the 3 ideas from this article I chose. Not just the polished final product, but also the rejected ideas, the wasted credits, and my face on the night before the deadline.

By the way, this is my representative Japanese-style MV. “Strike the Path” | Japanese Dark Fantasy MV https://www.youtube.com/watch?v=AxIppl5ndig (YouTube @ukiyo_il)

I want to submit something that surpasses this. No, I will.

Contests are, in the end, places where rankings are assigned. There are days when you lose. Actually, you lose most of the time. My magazine is even named “Cutting into AI Contests, and Usually Getting Beaten Back” (lol).

But because there is a deadline, the work gets finished. Because there is a ranking, you can get serious. Only those who have done it for real can move on to the next step.

If you find even one of the 30 project ideas in this article that you think “I want to make this,” please take it. If my delusions become someone’s entry, that alone makes writing this article worth it. It would be great if this article becomes the trigger for you to think, “Maybe I’ll submit one.” Once you decide which one to enter, let me know in the comments. If there’s someone entering the same contest, I’ll arbitrarily designate you as a rival and support you with all my might.

Thanks✨ Ukiyo🎴