
We’re excited to introduce Intelligence. If you’re building frontier models or working on hard-to-verify evaluations, we’d love to talk.
✓What’s the limit? Creators of @designarena, @predictionbench, @socialsarena

We’re excited to introduce Intelligence. If you’re building frontier models or working on hard-to-verify evaluations, we’d love to talk.

Model latency significantly impacts user satisfaction. Surprisingly, @OpenAI's speed decreases noticeably during the work day, while @AnthropicAI's doesn't at all. Almost all benchmarks measure latency at test time, but few look at how API response time performs live on production. We looked at 200,000 LLM requests submitted by users on DesignArena, at every hour of the day, sent to each lab's standard API. OpenAl model speeds decrease significantly during the work day, fluctuating up to 16%. Even more interestingly, Anthropic's models don't seem to be affected at all.

The @Alibaba_Qwen team has made huge improvements in Office work - this model is thorough: In the promotion-interview template below, Qwen4-Max lists 2 failure stories while Kimi K3.1 pro tells a single professional failure Congrats to the team!

BREAKING: Qwen4-Max by @Alibaba_Qwen ranks #4 on Python-PPTX arena with an Elo rating of 1302 and #5 on Spreadsheet arena with an Elo rating of 1336. On Spreadsheet arena, this marks a 30-position jump from Qwen’s previous model, Qwen 3.7 Max, and a 379-point Elo increase. On Python-PPTX arena, this marks a 5-position jump from its previous model and an 80-point Elo increase. This also places it in the same performance band as top models from @AnthropicAI, @OpenAI, and @Kimi_Moonshot. Congrats to the @Alibaba_Qwen team on placing top-5 on both leaderboards!

Introducing LifeArena: the world’s first arena for personal assistants, as judged by real users. As AI assistants become more capable, existing static benchmarks struggle to capture the creative, unthinkable requests of people in their everyday lives. So, we offered select users the most popular personal assistants: @instinct, @Muse by @AIatMeta, Dots by @OpenAI, Grok Bot by @xai, and Gemini by @GoogleDeepMind. The results are in: @Muse is #1... by a landslide. Here's what users said:

The majority of the human experience is visual. Proud to see @DesignArena representing the visual coding capabilities in @thinkymachines' model card. Excited to see visual coding become an increasingly dominant frontier for AI evaluation!

BREAKING: Inkling 2 Small by @thinkymachines takes 4th place overall on DesignArena and is 1st place among open weights models with an Elo of 1376! At $0.30/$1.00 per 1M input/output tokens, Inkling 2 Small is 15× cheaper on output than Kimi K3 by @Kimi_Moonshot. This model also places 1st on 2 of our subcategories including General Purpose and Landing Page prompts. Model weights to be released soon. A huge congrats to the @thinkymachines team on a top-5 placement on the leaderboard!


Your model loves making up fake ratings to make your business look good. Their favorite rating is 4.9 stars. This page came from a 6-word request with no name, numbers or ratings. Claude Fable 5 invented the stars, # reviews, # new styles, free 30-day returns and even the live amount of customers. Across ~7k websites AI models built on DesignArena, 451 of the 681 with a star rating say 4.9.

Introducing Code to Video Arena. Models have gotten remarkably good at creating animated videos without using video models at all. Instead of generating frames through diffusion, they can write HTML, CSS, and JavaScript that render into a film. Two models receive the same prompt, and each directs a 15–45 second film entirely through code, contained in a single webpage. We render both films at 1080p and 60fps. Each film also gets its own sound design, synthesized through the Web Audio API, and is normalized to the same loudness. Leaderboard coming soon! Try it today at designarena.ai

Models can seemingly degrade over time. The following scenes were created with the same prompt and the same model, @GoogleAI's Gemini 3.8 Flash, by a user on DesignArena. The first was generated just 3 days after release, while the second came almost a month later. The difference is striking. The first scene features a finished bungalow with terraced gardens, while the second is barely recognizable. Digging into the code, we find that the second scene still attempts to build the garden, windows, and patio, but buries much of it beneath its own terraces. In aggregate, the share of objects never visible on camera jumps from 18% to 57%. For bushes, the share placed underground jumps from 0% to 100%, and even the house walls go from fully above ground to 64% below it. Both scenes share the same simple user request, yet a few weeks after release, the model struggles to fulfill it. Try making your own generations and judge for yourself at DesignArena.ai!

BREAKING: Kimi K3.1 Pro by @Kimi_Moonshot is projected to rank #1 among open-weights models on DesignArena's Website leaderboard, and is tied for #3 among all models. Kimi K3.1 Pro has a 15 Elo point lead on their previous model, Kimi K3, on the open weights leaderboard. Among all models it ranks behind only Muse Wonder (max) and Muse Spark 1.3 (xhigh). That ties it with Muse Spark 1.3 Max for third place on the overall Website leaderboard. Congrats to the @Kimi_Moonshot team on the launch!

Canva’s first-ever image model has debuted at #4 on @DesignArena This makes @canva the #2 AI image lab in the world, behind only @OpenAI An incredible achievement from a very hardworking and talented team pushing the boundaries of what design companies accomplish in frontier AI research Huge congratulations to @MelanieCanva and the rest of the @canva team!

Canva Craft, the first ever image model by @canva, takes the #4 spot on DesignArena's Image leaderboard. This makes Canva the #2 Image Lab after @OpenAI! Notably, it wins most often on photorealistic product and object shots, setting up studio-style lighting and detailed material textures on a single hero object. It also ranks #1 on the Portrait and Landscape subcategories. Congrats to the @canva team on the launch!

BREAKING: FLUX 3 Max by @fal lands at #6 in Video and Image-to-Video on Design Arena, debuting in the top 10 for both leaderboards. This is the highest fal model yet on the Video leaderboard. This also places @fal's new model alongside leading video models from @Alibaba_Wan, @bfl_ai, and @GoogleDeepMind. Congrats to the @fal team on the launch!

BREAKING: HeyGen Video 1.0 (1080p, ultraquality) by @HeyGen, a post-trained model on MiniMax H3 by @MiniMax_AI, ranks #3 on both Design Arena's Video and Image to Video leaderboards, making HeyGen the #2 Video lab overall! In Video, HeyGen Video 1.0 sits just behind Gemini Omni Flash by @GoogleDeepMind and ahead of Wan 3.0 by @Alibaba_Wan. In Image to Video, it also sits just behind MiniMax H3 Max by @MiniMax_AI, while generating faster than 71% of models. Congrats to the @HeyGen team!