Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Fable 5.1 one-shot this Dark Souls meets Diablo game. All code, 222 KB total, zero asset files. Every texture, model and sound effect is generated procedurally. The whole sfx system is 7 KB of Web Audio synthesis. It nailed mechanics I only named in text: stamina, dodge rolls, barbarian...

146,907 Aufrufe • vor 7 Tagen •via X (Twitter)

60 Kommentare

Profilbild von Dave Senpai
Dave Senpaivor 7 Tagen

It's going to be a thing. I'm building an web MMO also, 3km+ world with hundreds of detailed assets and a total size of 2mb (637kb gzipped). Loads in under 20 seconds. Twilight zone stuff.

Profilbild von Meng To
Meng Tovor 7 Tagen

Crazy. Can’t wait to take this to the next level

Profilbild von wbk ᕦ(ò_ó )ᕤ
wbk ᕦ(ò_ó )ᕤvor 7 Tagen

That is cool! How long did it take? My Souls-like tests I made a while ago pre-Fable5.1:

Profilbild von Meng To
Meng Tovor 7 Tagen

Nice! 5.1 works way faster

Profilbild von wbk ᕦ(ò_ó )ᕤ
wbk ᕦ(ò_ó )ᕤvor 7 Tagen

My first 5.1 test went a little wrong lol

Profilbild von Fluxora
Fluxoravor 7 Tagen

That's mad altogether ! the quality is impressive how long did it take ?

Profilbild von Meng To
Meng Tovor 7 Tagen

A bit more than 3 hours

Profilbild von Gene
Genevor 7 Tagen

I don’t understand how… I tell it things and it doesn’t produce anything close to this 😂 Are you really specific on the prompt or vague?

Profilbild von Bryant
Bryantvor 7 Tagen

Well done! will you keep building this? open source?

Profilbild von Meng To
Meng Tovor 7 Tagen

yep, I'll keep building

Profilbild von Bryant
Bryantvor 6 Tagen

let me know if you need a beta tester!

Profilbild von MARUF
MARUFvor 7 Tagen

procedural web audio synthesis in 7 kb is crazy impressive

Profilbild von Meng To
Meng Tovor 7 Tagen

Let’s see how quality scales!

Profilbild von Webster | JARVIS
Webster | JARVISvor 7 Tagen

this is wild. 222kb and zero assets, feels like cheating lol procedural audio is slept on, 7kb for a whole sfx system is insane

Profilbild von Jacqueline van Wyk
Jacqueline van Wykvor 6 Tagen

222kb for a whole game is unreasonable honestly. my photo library alone is bigger than this man's entire boss fight

Profilbild von Groov Mekanik
Groov Mekanikvor 7 Tagen

Nice. Sound is terrible haha but thats good for us audio engineers, a good sound designer that knows what they’re doing can use web assembly to make something 1000x better and it still be efficient.

Profilbild von Lê Anh Ngọc
Lê Anh Ngọcvor 7 Tagen

could you share the link to try it

Profilbild von Meng To
Meng Tovor 7 Tagen

Will make it available

Profilbild von Technomantics™ - °A Shrewd Skew° heading to Steam
Technomantics™ - °A Shrewd Skew° heading to Steamvor 6 Tagen

I wish more people understood most AI tools are repurposing existing templates, blueprints, pre-made games, & porting them while updating to some of your asks. Create actually new, novel, innovative mechanics, gameplay, new approach to physics/collisions/animations//etc; handling of anything outside the norm, & they all have to figure it out or have tons of bugs while you go back & forth with them. Meanwhile, large AI firms make bank for users being gullible.

Profilbild von George Papadakis
George Papadakisvor 7 Tagen

Make it an MMO and release it :)

Profilbild von Meng To
Meng Tovor 7 Tagen

graphics isn't perfect, but we can get there! no 50gb download. :)

Profilbild von George Papadakis
George Papadakisvor 7 Tagen

I wonder what would Astra or Fable 5.1 do if u would ask them to just "Make it a professional fully functional MMO with chat facility and release it". I would try it, but I am still waiting for Astra and Fable does not last for a day these days :)

Profilbild von Meng To
Meng Tovor 7 Tagen

astra is probably as expensive as fable 5.1. but openai is more generous and the resets help.

Profilbild von George Papadakis
George Papadakisvor 7 Tagen

@thsottiaux for president :)

Profilbild von Dr.BMW
Dr.BMWvor 6 Tagen

Aaaand it's obsolete

Profilbild von Tesana - Make games with AI
Tesana - Make games with AIvor 7 Tagen

Looks Great!

Profilbild von Meng To
Meng Tovor 7 Tagen

my last attempt was much worse:

Profilbild von 燎原 | 独立开发者
燎原 | 独立开发者vor 7 Tagen

那我很好奇,既然没有任何美术参与,你的关键词是怎么样的

Profilbild von Pizza Yap 🍕
Pizza Yap 🍕vor 7 Tagen

Let’s team up and conquer! 🔥

Profilbild von Meng To
Meng Tovor 7 Tagen

triple a browser games need to happen.

Profilbild von Pizza Yap 🍕
Pizza Yap 🍕vor 7 Tagen

Woah 3A browser game! Yes agree this need to be happen!

Profilbild von Ryan Edkins
Ryan Edkinsvor 7 Tagen

222KB of pure code with zero asset files, and it still nailed dodge rolls and boss mechanics you only described in text, genuinely impressive work.

Profilbild von Barış AKIN
Barış AKINvor 6 Tagen

how can it be one-shot with dozen prompts

Profilbild von Meng To
Meng Tovor 6 Tagen

That was just asking how big the html is.

Profilbild von Greg Pazo
Greg Pazovor 6 Tagen

This is so rad!

Profilbild von insodimension
insodimensionvor 6 Tagen

i played this back in the day opus 5 launch. fun to play. PO2/Diablo style.

Profilbild von Julien Tavernier
Julien Taverniervor 6 Tagen

Prompt ? 🙏🏼

Profilbild von InfoRemix
InfoRemixvor 7 Tagen

222kb 🤯 nimble procedural wonder

Profilbild von Ved 💫
Ved 💫vor 7 Tagen

Curious...how much did this cost ?

Profilbild von Meng To
Meng Tovor 7 Tagen

probably less than $5, worked for a few hours.

Profilbild von Ved 💫
Ved 💫vor 7 Tagen

Crazy ! no api calls just straight up claude code ?

Profilbild von Eamin
Eaminvor 7 Tagen

Direction becomes the skill

Profilbild von Tom Lee
Tom Leevor 7 Tagen

So awesome. Where is this playable?

Profilbild von Meng To
Meng Tovor 7 Tagen

I’ll upload it soon

Profilbild von Bexel Initiative - Let's Talk Amendments!
Bexel Initiative - Let's Talk Amendments!vor 7 Tagen

Zero assets? 🤔🤔🤔

Profilbild von Meng To
Meng Tovor 7 Tagen

all code!

Profilbild von Bexel Initiative - Let's Talk Amendments!
Bexel Initiative - Let's Talk Amendments!vor 7 Tagen

Cray! Wow.

Profilbild von ✌︎ frederik ✌︎
✌︎ frederik ✌︎vor 7 Tagen

have you tried the new gpt 6 with this flow yet?

Profilbild von Meng To
Meng Tovor 7 Tagen

Just waiting for my access

Profilbild von ✌︎ frederik ✌︎
✌︎ frederik ✌︎vor 7 Tagen

me too

Profilbild von Abandoned Films
Abandoned Filmsvor 6 Tagen

Share the game then.

Profilbild von Meng To
Meng Tovor 6 Tagen

Doing a huuuge upgrade before sharing

Profilbild von SamsonCJ
SamsonCJvor 7 Tagen

222KB!!?

Profilbild von Abeansits
Abeansitsvor 6 Tagen

you forgot "FRESH MEAT!"

Profilbild von Aeris Vahn Ephelia - 𝑹𝑬𝑽
Aeris Vahn Ephelia - 𝑹𝑬𝑽vor 7 Tagen

Sounds really awesome. I want to generate this

Profilbild von sezereth
sezerethvor 2 Tagen

So now that everyone can create "games" with a push of a button, who will sort through all the crap to find the good stuff?

Profilbild von Meng To
Meng Tovor 2 Tagen

this isn't just clicking one button. the baseline will just keep going higher, always requiring more.

Profilbild von Shokunin - Product Craftsmanship
Shokunin - Product Craftsmanshipvor 7 Tagen

Eeeeeeeh. I mean it's good content. It's a nice video and I'm sure it's fun to build. Very far from a game though

Profilbild von Bad Major
Bad Majorvor 6 Tagen

what about time and $?

Profilbild von Legion Forged Gaming
Legion Forged Gamingvor 6 Tagen

Prototypes are not "one-shots"

Ähnliche Videos

fable 5.1 vs fable 5 vs opus 5 – three lord of the rings landmarks, built in 3d from one image the setup: one reference image per scene, one html file per build, everything procedural – no meshes, no textures, no image files, nothing past Three.js from a cdn. each model reads the picture, writes its own prompt from it, then builds to that prompt in the same turn. three named camera shots per scene on keys 1/2/3, so it can be screen-recorded. run through OpenRouter tasks: 1. bag end – hobbiton from two frames, outside and in. the round green door has to open onto the room you are standing in 2. barad-dûr – the tower and orodruin from one film still. the eye has to move and track the camera, the volcano erupts on a cycle, the clouds never stop 3. rivendell – jerry vanderstelt's painting. sun shafts that shimmer, water that falls without a break, trees that sway on a gust models: Anthropic fable 5.1, fable 5, opus 5 total cost, three builds #1 fable 5 – $14.97 #2 opus 5 – $18.53 #3 fable 5.1 – $22.38 wall clock, three builds #1 fable 5 – 38m #2 fable 5.1 – 92m #3 opus 5 – 122m output tokens #1 fable 5 – 298,592 #2 fable 5.1 – 439,435 #3 opus 5 – 724,418 lines of code shipped #1 fable 5 – 2,885 #2 fable 5.1 – 4,021 #3 opus 5 – 5,161 biggest single build, lines #1 opus 5, bag end – 2,410 #2 fable 5.1, barad-dûr – 1,375 #3 fable 5, bag end – 1,319 observations: • fable 5.1 is the only model that furnished the bag end interior – a live fire, panelling, books on the floor, leaded diamond windows, against fable 5's flat color and opus's dark tunnel. the round door outside opens onto that room, the hard part of the brief • what it costs is thinking room. the 128k output ceiling is a thinking budget in disguise: fable 5.1 burned 102,116 of it on reasoning and hit the wall mid-file. opus spent 109,241 and hit the same wall. fable 5 spent 61,240 and finished bag end in one call – the only one that did • fable 5.1's first pass is not the finished thing. its barad-dûr came back with three defects you only catch by looking at it – nothing a read of the code would have flagged • it is the best of the three at being corrected. handed a plain list of what was wrong, it returned 32 targeted patches over two rounds, every one applied first try, and it worked out one of the causes itself instead of guessing at constants conclusion: nine scenes, 12,067 lines and 1.46m output tokens for $55.88 all in – and the cheapest model was also the fastest, by 3.2x! follow thehype. for 24/7 ai news, analysis and breakdowns

thehype.

18,509 Aufrufe • vor 8 Tagen

Fable 5 just turned me into an iOS developer. I spent 2 days building a full game with zero code written by hand. The last 10 builds succeeded on the first try. I needed a fun project to test it on, so I decided to make an iOS game. After a few experiments with other ideas, I settled on a runner starring Dario as the main character, called "Credit Runner". Hilarious, I know! In terms of credits spent, my Claude Max plan stands at 36% right now, and this isn't the only project I ran through Fable 5. A few quick notes: - This is just an early prototype, a long way from release, but it's really fun to play! - It's extremely fun to work with Fable 5! - Fable 5 wants to get things done, and it pushes you towards completion much faster than any other model. - Once we established the ground rules, how I wanted the game built, and the overall structure, the last 10 builds were flawless, succeeding on the first try. This is impressive! - All the assets were generated with AI (Magnific and Trippo for the visuals, ElevenLabs for the voices). The only thing that's not AI is the music. I just really liked this track. - There are many rules in the game designed to make it really fun to play, especially the randomness. - There are a lot of visual cues inspired by San Francisco, and I hope you catch them all. - At some point on the first day, I honestly thought I wasn't going to get where I wanted, but Fable pushed me towards the goal, and by the second day it started to deeply understand the project. - The prompts were really long, especially in the beginning, to make sure everything was covered. - Once the app was solid, I tried to pack as many changes as I could into a single prompt to maximize my Fable usage. This was just a fun project, but if anyone wants to play it, I might put the final, much more polished version on the App Store. What do you think? Should I?

Alex Patrascu

11,824 Aufrufe • vor 3 Monaten

This is probably the most complex workflow I’ve ever built, only with open-source tools. It took my 4 days. It takes four inputs: author, title, and style; and generates a full visual animated story in one click in ComfyUI . I worked on it for four days. There are still some bugs, but here’s the first preview. Here’s a quick breakdown: - The four inputs are sent to LLMs with precise instructions to generate: first, prompts for images and image modifications; second, prompts for animations; third, prompts for generating music. - All voices are generated from the text and timed precisely, as they determine the length of each animation segment. - The first image and video are generated to serve as the title, but also as the guide for all other images created for the video. - Titles and subtitles are also added automatically in Comfy. - I also developed a lot of custom nodes for minor frame calculations, mostly to match audio and video. - The full system is a large loop that, for each line of text, generates an image and then a video from that image. The loop was the hardest part to build in this workflow, so it can process either a 20-second video or a 2-minute video with the same input. - There are multiple combinations of LLMs that try to understand the text in the best way to provide the best prompts for images and video. - The final video is assembled entirely within ComfyUI. - The music is generated based on the LLM output and matches the exact timing of the full animation. - Done! For reference, this workflow uses a lot of models and only works on an RTX 6000 Pro with plenty of RAM. My goal is not to replace humans, as I’ll try to explain later, this workflow is highly controlled and can be adapted or reworked at any point by real artists! My aim was to create a tool that can animate text in one go, allowing the AI some freedom while keeping a strict flow. I don’t know yet how I’ll share this workflow with people, I still need to polish it properly, but maybe through Patreon. Anyway, I hope you enjoy my research, and let’s always keep pushing further! :)

Lovis Odin

58,841 Aufrufe • vor 11 Monaten