Design Arena (@DesignArena) — posts, photos & videos

@DesignArena

Oct 7

Claude Haiku 5.5 by @AnthropicAI is now available on Design Arena!

Built for high-volume, cost-sensitive tasks, Claude Haiku 5.5 is Anthropic’s fastest and most capable small model yet, with support for coding, classification, summarization, database queries, and speed-sensitive workflows like customer support and browser use. It costs around 75% less to run than Haiku 4.5, making it well suited for efficient, large-scale workloads.

Congrats to the @AnthropicAI team on the launch!

Claude@claudeai · Oct 7

Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released.

On average, it costs around 75% less to run than Claude Haiku 4.5.

0:11

6 3 0 84 3K 2

@DesignArena

Oct 7

Since Opus 5.5 by @AnthropicAI was released just over 2 weeks ago, it has become a fan-favorite amongst developers, designers, and casual users alike.

It has taken 1st place on 4 of Design Arena’s leaderboards (Overall Frontend, Data Visualization, 3D Design, and React Native). Additionally, it is Top 3 on our Game Dev and UI Components leaderboards.

We noticed a couple of reasons why this may be the case:

17 33 6 553 24.7K 67

@DesignArena

Oct 6

Nano Banana 2.1 by @GoogleDeepMind is now available on Design Arena!

The latest image generation and editing model from Google brings improvements across visual design, mask-based editing, and subject consistency. Nano Banana 2.1 is built to create more natural-looking images with greater control across editing workflows.

Congrats to the @GoogleDeepMind team on the launch!

Google@Google · Oct 6

Meet Nano Banana 2.1, our latest image generation and editing model 🍌

This upgraded version outperforms our previous models across the board, with notable leaps in visual design, mask-based editing, and subject consistency to help you create more natural-looking images.

GIF

4 2 0 55 2.9K 4

@DesignArena

Oct 4

BREAKING: GPT-6 Astra by @OpenAI has taken the #1 spot across 4 of Design Arena’s leaderboards: 3D Design (1484), Frontend (1397), Full Stack (1355), and Image-to-HTML (1272).

This is an impressive takeover since its release just a month ago - in particular, it excels at 3D Design (see example below):

108 79 74 1.1K 148K 272

@DesignArena reposted

Alex Gong

@gongzhongcheng

Oct 4

We found in the thinking summaries that Opus 5.5 structures its thinking and executes more like a "builder", while Astra approaches problems more like a "designer". Go check out the new video now!

Design Arena@DesignArena · Oct 2

@OpenAI's GPT-6 Astra hedges ("maybe", "might", "it seems") about 20x as often as @AnthropicAI's Claude Opus 5.5. Opus usually weighs a few options and commits early, in about 4 out of 5 summaries as opposed to Astra's 1 in 4.
We read 324 of their thinking summaries on Design Arena to see how each one actually goes about building a game. Astra works more like a designer, Opus more like a builder.
Full video in the replies.

0:27

4 5 0 17 1.3K 7

@DesignArena

Oct 2

0:27

14 13 3 142 18.3K 27

@DesignArena

Oct 2

FLUX 3 Image by @bfl_ai is now available on Design Arena!

FLUX 3 Image brings precise pixel-level changes with multi-turn editing, bounding-box control, and support for up to 10 reference images. It can generate images up to 4K while preserving detail, with commercial weights available for companies running image generation at scale.

Congrats to the @bfl_ai team on the launch!

Black Forest Labs@bfl_ai · Oct 1

Introducing FLUX 3 Image.

Control every pixel.

Make precise multi-turn edits without changing any other pixel.

Lay out the image exactly how you want using bounding boxes.

Generate in up to 4K to preserve details.

Use up to 10 references to compose an image.

Commercial Weights available for companies running image generation at scale.

Open Weights version of FLUX 3 Image is launching in the coming weeks.

0:49

7 2 1 43 2.6K 8

@DesignArena

Oct 2

BREAKING: P-Video-2-Pro (Cost) by @PrunaAI lands on the Pareto frontier for Preference vs Price in both Video and Image to Video on Design Arena.

P-Video-2-Pro (Cost) is 3× cheaper than P-Video-2-Pro (Quality) at only $0.125 per 5-second video, while having an average generation time of ~10 seconds, nearly 10× faster than the average video model.

Congratulations to the @PrunaAI team on this achievement!

7 10 1 61 3.9K 9

@DesignArena

Sep 30

BREAKING: Ling-3.1-flash by @AntLingAGI debuts as the #2 open-weights model on Mobile App Arena by Design Arena with an Elo of 1207.

Ranking #17 overall with performance comparable to Opus 4.6, Ling-3.1-flash performs particularly well at building mobile productivity trackers and games.

Congratulations to the @AntLingAGI team on this achievement!

6 18 2 198 14.1K 39

@DesignArena

Sep 30

BREAKING: Recraft V4.1 Flash by @recraftai is the fastest image model yet on Design Arena.

With an average generation time of just 1.8 seconds, 16× faster than the average model, it overtakes P-Image by @PrunaAI to claim the top spot for speed and lands on the Image Pareto frontier for Preference vs Speed.

Congratulations to the @recraftai team on this achievement!

3 3 1 52 4.2K 12

@DesignArena

Sep 30

BREAKING: Ideogram 4.5 by @ideogram_ai places 15th in Image Editing with an Elo of 1250.

This puts the model in the same performance band as MAI-Image-2.6 by @MicrosoftAI and Gemini 3 Pro Image Preview by @GoogleDeepMind. The model performs particularly well with typography edits, such as modifying text in infographics.

Congratulations to the @ideogram_ai team on the launch!

23 10 3 153 24.2K 32

@DesignArena

Sep 29

GPT-6.1 Sol by @OpenAI is now available on Design Arena!

An upgrade to GPT-6 Sol, GPT-6.1 Sol brings near-GPT-6 Astra intelligence across agentic coding, computer use, and professional work at one-fifth of Astra’s standard token prices. Cached input costs just $0.10 per million tokens, making it more cost-effective to build and run capable agents that reuse context across requests.

Congrats to the @OpenAI team on the launch!

OpenAI@OpenAI · Sep 29

GPT-6.1 Sol: near-Astra intelligence for a fifth of the price.

It’s the most cost-efficient model for its performance available today.

0:08

2 2 0 38 2.9K 1

@DesignArena

Claude Sonnet 5.5 by @AnthropicAI is now available on Design Arena!

The second model in the Claude 5.5 family, Sonnet 5.5 is a clear upgrade over Sonnet 5, with stronger performance across coding and everyday knowledge work. It runs more than 30% faster and costs up to 30% less for most tasks, while using fewer tokens to get the same work done. It also brings stronger design capabilities for creating polished documents, slides, spreadsheets, and user interfaces.

Congrats to the @AnthropicAI team on the launch!

Claude@claudeai · Sep 28

Introducing Claude Sonnet 5.5, the second model in the Claude 5.5 family.

It’s a clear upgrade over Sonnet 5, runs more than 30% faster, and costs up to 30% less for most work.

0:13

4 2 0 70 3.1K 2

@DesignArena

BREAKING: Arrow 2 Telos by @QuiverAI ranks #1 on SVG Arena by Design Arena with an Elo of 1624. It's the first model to ever break 1600+ on our SVG leaderboard, establishing the new SOTA frontier for SVG generation.

Massive congrats to the @QuiverAI team for this achievement!

12 44 18 478 94K 298

@DesignArena

This palm tree is floating 5 units above the ground.
@OpenAi's GPT-6 Astra and @Xiaomi Mimo 2.6 both built a fictional pacific island. Astra read the terrain before planting trees, while Mimo placed a floating tree in 73% of it's generations.

We dove into the code to see what makes an AI-generated world actually work. Try it for free on DesignArena!

0:21

3 5 0 40 3.3K 6

@DesignArena

P-Video-2 Pro (Quality) and P-Video-2 Pro (Speed) by @PrunaAI tie each other for #2 on Design Arena's Image to Video leaderboard with an Elo of 1325.

Based on MiniMax H3 by @MiniMax, P-Video-2 Pro (Quality) generates in 8.0 seconds, while P-Video-2 Pro (Speed) has an 4.5-second generation time, placing it on the Pareto frontier of Preference vs Speed.

Congratulations to the entire @PrunaAI team!

7 13 5 64 14.6K 11

@DesignArena

Sep 26

We were surprised to find that 50% of Open AI's GPT-6 Astra's driving games have mistakenly reversed steering mechanics.

So what makes users prefer games, and where do the models fall short? We analyzed 4,300+ tournaments in DesignArena to find out.

Beyond looks, there are other interesting axes to explore: whether all scene objects obey the same physics, whether the controls work the way players expect, and whether the world interactions hold up once you start moving.

Full video in the thread. Try it for free on DesignArena!

0:09

11 7 6 151 21K 49

@DesignArena

Sep 23

Recraft V4.1 Flash by @recraftai is now available on Design Arena!

Built for speed, Recraft V4.1 Flash generates images in as little as 1.3 seconds, keeping creative workflows moving through rapid iteration and variations. At a fraction of the cost of V4.1, it brings the same Recraft look across photography, portraits, complex scenes, mood and atmosphere, and early concepts.

Congrats to the @recraftai team on the launch!

Recraft@recraftai · Sep 23

The fastest image model on the market is here.

Zero lag in your workflow.

Recraft V4.1 Flash produces images in 1.3 seconds.
Pick Flash from the model picker, generate, then hit Refine to clean up the details.

Check it out and give it a try ↓

0:33

2 2 0 41 2.5K 3

@DesignArena

Sep 22

GPT-6 Sol & GPT-6 Luna by @OpenAI are now available on Design Arena!

Built with the same advances behind GPT-6 Astra, Sol and Luna bring improved performance across professional work, factuality, coding, computer use, and alignment in faster, more affordable models. With improved caching and inference, both models are designed to make advanced AI more accessible for everyday tasks and applications at scale.

Congrats to the @OpenAI team on the launch!

OpenAI@OpenAI · Sep 22

Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe.

GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale.

We’ve also made caching and inference more efficient, and we’re passing the savings directly to you: 50% lower API prices for Sol and Luna compared with GPT‑5.6 promotional pricing.

0:08