Prompt no longer exists? DALL·E3 suddenly merged into ChatGPT, with accurate drawing and realistic details. OpenAI really blew up the AI ​​painting circle today. When I woke up in the morning, the AI ​​painting circle had been blown away. OpenAI dropped a blockbuster - DALL·E3, which merged with ChatGPT, was officially announced!



One is the most powerful language model in history, and the other is the most powerful Vincentian graph model in history. The combination of the two is bound to subvert the entire AIGC circle.

After undergoing the most powerful transformation, DALL·E3 can understand the nuances of semantics, and the details are terrifyingly realistic.


Prompt: "Illustration of an avocado sitting in a therapist's chair saying 'Ijustfeelsoemptyinside' with a crater sized hole in the middle. Therapist, a spoon, scribbling notes"

Netizens exclaimed: "If DALL·E3 can really accurately represent the text in the image, then it will completely change the rules of the game!!!"


At the same time, the threshold for prompts has also been greatly reduced.

Users who "die early" no longer have to worry about not being able to write gorgeous prompt words. Just give me a hint and ChatGPT will write eloquent paragraphs for you.

In short, you can let your imagination run wild and let ChatGPT take you flying!


No more racking your brains to figure out how to prompt! Just by talking to ChatGPT, you can write wonderful things

You can even make changes at any time based on the generated image during the conversation.

It can be seen that in order to catch up with the release of multi-modal Gemini by Google, OpenAI took the lead.

Jim Fan, senior AI scientist at Nvidia, said excitedly that this is a preview of the epic battle between large-scale multi-modal LLM and Google Gemini!

OpenAI CEO Sam Altman also said proudly in the comment area: "Wearesoback".


Currently, DALL·E3 is still in internal testing and is expected to be launched to all Plus users and enterprise users in early October.

Midjourney’s rival is really here!

DALL·E3 is unique in that it is built natively on ChatGPT.

Just make a request to ChatGPT, a simple sentence, or a detailed paragraph, and it will start brainstorming and improving tips.


Nvidia's top scientist Jim Fan said, this is so shocking!

The significance of DALL·E3 is that it is not only a strong opponent of Midjourney,It also represents that the war of multi-modal large models is about to break out.

DALL·E3's excellent language matching capabilities are based on a solid text GPT foundation. Midjourney doesn't really have much of a "reasoning brain," which is why so many hint hacks are needed.

Brains first, pixels later, this is how to build powerful multi-modal AI.


There is no need to learn prompt engineering, you can just do whatever you want. For prompt words, ChatGPT will help you write it, and the details of the exquisite paintings will be taken care of by DALL·E3.

The following example shows you how DALL·E3 meticulously engraved every scene in the text into the painting when generating the following painting, and the picture quality is like a movie.

Under a full moon, the streets were filled with bustling pedestrians enjoying the nightlife.

At a corner stall, a girl with fiery red hair and an elegant velvet cloak was haggling with the grumpy boss.

The boss, tall and burly, wearing a crisp suit and a striking goatee, was chatting animatedly on his retro steampunk-style phone.


In addition, under the same prompt, the image drawn by DALL·E3 will be more accurate than that of DALL·E2.

For example, "An expressive oil painting of basketball player dunking, depicting asan explosion of anebula", DALL·E3 is more detailed in the generation of characters, basketballs, and more macro backgrounds.


As can be seen from the comparison chart made by netizens, the gap between DALL·E3 and DALL·E2 can be said to be epic.


Netizens said that OpenAI is not just trying to surpass Midjourney this time. Google is their biggest competitor, especially recently Gemini has been imminent.


Some people say that this change reflects an important evolutionary trend in AI painting tools:understand people.

It would be great if AI allowed us to better describe and modify images! Let us wait and see how far DALL·E3 can go.


DALL·E3+ChatGPT

DALL·E3+ChatGPT, how amazing is it?

Altman hid the cute details generated by DALL·E3 in a promotional video.


DALL·E3 can turn a 5-year-old child's fantasy into reality.

"My 5-year-old baby has been talking about a "super sunflower hedgehog". What should it look like? "

ChatGPT automatically generates images of different styles based on 4 different prompt words. By selecting the last one here, you can set the base picture for continued creation.


Next, name the little hedgehog Larry and draw a house for him.

It can be seen that DALL·E3 not only accurately understood the requirements, but was also able to annotate text in the picture, including "Larry" on a mailbox.

Completely solved the problem of the previous DALL·E model not being able to write at all.


You can also complete the storyline while drawing, and even add illustrations to it.


In addition, DALL·E3 can also generate cute stickers based on the above styles.


Finally, let ChatGPT summarize the above content and generate a bedtime story, which is absolutely perfect.


OpenAI official also provides a large number of exquisite photo albums.

prompt: This is a paper art piece depicting a girl gently hugging a cat. They were both sitting in a potted plant, the cat purring contentedly and the girl smiling. The scene is dotted with handmade paper flowers and leaves.


Apapercraftartdepictingagirlgivinghercatagentlehug.Bothsita amidpottedplants,withthecatpurringcontentedlywhilethegirlsmiles.Thesceneisadornedwithhandcraftedpaperflowersandleaves.

DALL·E3 can create beautiful posters and add text.

prompt: Venus retro travel poster, portrait. The image depicts Venus surrounded by thick yellowish clouds and the silhouette of an old rocket ship approaching. Mysterious shapes suggest mountains and valleys beneath the clouds. The text at the bottom reads "Exploring Venus: The Beauty Behind the Fog." The color scheme consists of gold, yellow and soft orange, evoking curiosity.


AvintagetravelposterforVenusinportraitorientation.Thesceneportraysthethick,yellowishcloudsofVenuswithasilhouetteofavintagerocketshipapproaching.Mysteriousshapeshint atmountainsandvalleysbelowtheclouds.Thebottomtextreads,'ExploreVenus:BeautyBehindtheMist'.Thecolorschemeconsistsofgolds,yellows,andsoftoranges,evokingasenseofwonder.

prompt: This illustration depicts a human heart made of translucent glass, standing on a pedestal amid stormy waves. A ray of sunlight penetrates the clouds, illuminating the heart and revealing the small universe within. On the horizon, the monument is engraved with a line of eye-catching characters: "Find the universe within you."


Anillustration of a human heart made of translucentglass, standingonapedestala amid a stormy sea. Rays of sunlight pierce the clouds, illuminating the heart, revealing atinyuniversewithin.

prompt: This is a 3D rendering of a coffee cup placed on the windowsill on a stormy day. The storm outside the window is reflected in the coffee cup, where miniature lightning and crashing waves can be seen. The interior is dimly lit, adding to the dramatic atmosphere.


A3Drenderofacoffeemugplacedonawindowsilduringastormyday.Thestormoutsidethewindowisreflectedinthecoffee,withminiaturelightningboltsandturbulentwavesseeninsidethemug.Theroomisdimlylit,addingtothedramaticatmosphere.

prompt: A detailed oil painting of an old captain steering his ship through a storm. Salt water splashed on his weathered face, his eyes were determined. Above, swirling evil clouds can be seen, harsh waves threaten to swamp the ship, while seagulls dive and swirl across the chaotic landscape. Thunder and lights were heard in the distance, illuminating the scene with an eerie green glow.


Adetailedoilpaintingofanoldseacaptain,steeringhisshipthroughastorm.Saltwaterissplashingagainsthisweatheredface,determinationinhiseves.Twirlingmalevolentcloudsareseenaboveanosternwavesthreatentosubmergetheshipwhileseagullsdiveandtwirlthrouahthechaoticandscape.Thunderandlightsembarkinthedistance.illuminatingthescenewithaneeriegreenglow.

Midjourney is so weak

As soon as the news came out, a wave of comparison pictures with Midjourney quickly appeared on the Internet.

The same reminder, Midjourney and DALL·E3 are simply not on the same level.


Also, SDXL and DALL·E3 generate comparison charts.


Some netizens said that they used RawMidjourney, Non-rawMidjourney and DALL·E3 to generate pictures using the same prompt.


Some netizens have made more pictures of Midjourney and DALL·E3PK.












Brand new prototype, codenamed "Gizmo"

At the same time, netizen "TiborBlaho" also broke the news that OpenAI is developing a new prototype codenamed "Gizmo" for ChatGPT——

It has an updated UI interface and introduces a powerful "Gizmo Editor" for enterprise customers, allowing users to create their own "GPT".






Specifically, you can define for your GPT the profile picture, model (GPT-3.5 or GPT-4), capabilities (including DALL·E, web browsing, coding sandbox, plug-ins, etc.), enabled plug-in list, behavior (what this GPT is, what it does, what it should know), welcome message (what it should say when a user starts a new chat with this GPT), and decide whether this GPT is private.