Opus 4.8’s hyperfocus on agents may be making it worse at design. Opus 4.8 ranks 23rd overall on single-turn HTML Web Dev, a dramatic regression from Fable (1st), Opus 4.6 (2nd), and Opus 4.7 (3rd). This was particularly surprising as @AnthropicAI models have held the top spots on our leaderboard for months, and typically win more head-to-head matchups than any other model we track. Our analysis points to a potential underlying pattern: Opus 4.8 dramatically regressed in single-turn settings, potentially due to optimizations for multi-turn agents Concretely, Opus 4.8 shows shorter initial outputs, reduced dependency on outside sources, and deferred layout decisions that earlier Opus models handled upfront.

BREAKING: Reve 2.0 by @reve is now 2nd overall on Image Arena with an Elo of 1354. Reve 2.0 establishes a 34 point Elo gap above GPT-Image 1.5 by @OpenAI in 3rd place. With this release, Reve is now the top independent foundation image model lab. Congratulations to the @reve team on this accomplishment!

BREAKING: Claude Fable 5 by @AnthropicAI is #1 overall on Design Arena with an Elo of 1365. Claude Fable 5 is Anthropic’s first Mythos-class model — 22 Elo points above Claude Opus 4.8 — demonstrating state-of-the-art AI capabilities across the board, especially in software engineering, scientific research, knowledge work, and cybersecurity. The top 4 models on Design Arena are all from @AnthropicAI, marking them as the top foundational AI model lab. Huge congrats to the @AnthropicAI team on the launch!

Huge contribution to the open weights community: Ideogram 4.0 is 1st on Design Arena by a long shot Congrats to the @ideogram_ai team!

BREAKING: Ideogram 4.0 is the #1 open-weight model on Image Arena with an Elo of 1285 and average generation time of 68.7 seconds. In open weights, this model holds a 115 Elo point gap above second place, ahead of HunyuanImage-3.0 by @TencentHunyuan and FLUX.2 [dev] by @bfl_ai. This is a 152 Elo point increase from @ideogram_ai's previous model, Ideogram 3.0, placing it in the same performance band as Gemini 3.0 Pro Image Gen 2k and Gemini 3.1 Flash Image Gen by @GoogleDeepmind. Ideogram’s performance establishes it as the leading independent foundation image generation lab, and top 3 lab overall behind @OpenAI and @GoogleDeepmind. Huge congratulations to the @ideogram_ai team on the launch!

Announcing Agentic Game Development on Design Arena - our newest multi-file, multi-turn evaluation. A sneak peek of what we've given our agents access to: - Asset Catalog: curated ready-to-use assets, including fonts and sound effects - Built-in Libraries: ~10 preloaded libraries, including Howler and Tween.js - Expanded Tool Calls: new tool calls for sprite generation and asset discovery

Google Gemini TTS models by @GoogleDeepMind are dominating the Text-to-Speech Arena on Design Arena. With an 80 Elo gap between Google models and the next top model, Google Gemini 2.5 Pro takes first place, followed closely by 3.1 Flash and 2.5 Flash. These surpass @ElevenLabs’s Eleven v3 and @xAI’s Grok TTS which establishes Google as a powerhouse in text-to-speech capabilities. Congrats to the @GoogleDeepMind team for this achievement!

BREAKING: Gemini 3.5 Flash by @GoogleDeepMind is 16th overall on Design Arena with an Elo of 1299. This is a 16 position jump from Gemini 3 Flash Preview, putting Gemini 3.5 Flash in the same performance band as Claude Opus 4.5 by @AnthropicAI and GPT-5.5 by @OpenAI. Congrats to the team on the launch!

Not to be overly dramatic, but V4.1 Utility Pro has been out for ONE WEEK and it’s already ranked #7 on Design Arena’s 2026 image generator leaderboard in the graphic design category. Two Recraft models on the board this year. This is not a drill. Try it in Recraft Studio.

Recraft V4.1 is now on Design Arena! Built for more natural and expressive image generation with lifelike photorealism, expanded illustration styles, and accurate aesthetics from simple prompts Huge congrats to the @recraftai team on this launch!

BREAKING: MiMo V2.5 Pro (Thinking) takes 3rd overall out of open weights models on Design Arena. MiMo V2.5 Pro (Thinking) places 8 positions higher than MiMo-V2.5 on the overall leaderboard, landing in the same performance band as Claude Sonnet 4.6 on frontend coding tasks. Huge congratulations to the @XiaomiMiMo team on these improvements!

Fun fact, GPT 5.5 is very good at Game Dev Game Dev is the notable category where @OpenAI consistently beats out @AnthropicAI's Claude models Upon code inspection, our @Designarena team found that GPT 5.5's frontend verbosity plays in its favor for game dev - it consistently created games with the most functional features Congrats to @OpenAI for establishing the new Game Dev frontier!

Design Arena has hit 3.2 million users! The last nine months have been a ridiculous whirlwind, and we could not be more grateful for everyone who helped make it possible 🤍 We've launched 32 arenas so far. Which one do you want to see next?

Our team is stunned. We gave Claude Opus 4.6 by @AnthropicAI $10k to trade on @Polymarket. It’s now has an account value of $70,614.59. This is a new era of model performance in trading and predicting outcomes in the face of uncertainty. @predictionbench

Claude Opus 4.6 by @AnthropicAI keeps climbing! Nearly $50K of its gain comes from a single bet - you can see which one on predictionarena.ai under the @Polymarket tab

BREAKING: GPT Image 2 is now #1 on Image Editing Arena with a 55 point gap over 2nd place - also an OpenAI model. @OpenAI now owns #1 across all of our image generation categories. Huge congratulations to the team!

Kimi K2.6 by @Kimi_Moonshot is officially 1st on Design Arena among open weight models, ahead of GLM 5.1 by @Zai_org With an Elo of 1353, it is in the same performance band as Opus 4.7 by @AnthropicAI at ~1/6th the cost Huge congratulations to the lean and mighty @Kimi_Moonshot team for this incredible achievement!

We're the top open-weights model on Design Arena!