Spreadsheet Arena

Spreadsheets have entered the arena! ⚔️ Announcing Spreadsheet Arena, the first research platform for human preference rankings on LLM-generated spreadsheets. The results? @AnthropicAI Claude Opus is on top, but the gap is tighter than you’d think. w/ @LTIatCMU, @Cornell, and @scale_ai. 🧵

⚔️BREAKING: Gemini 3.1 Pro Preview debuts as #7 overall on Spreadsheet Arena, trailing Gemini 3 Pro by @GoogleDeepMind which currently stands at #6

⚔️BREAKING: Claude Sonnet 4.6 by @AnthropicAI debuts at #2 in Spreadsheet Arena, trailing Opus 4.6!

Sonnet 4.6 is now on Spreadsheet Arena! How well can it model Anthropic's Series G?

⚔️BREAKING: Claude Opus 4.6 by @AnthropicAI debuts at #1 in Spreadsheet Arena, surpassing Opus 4.5!

TL;DR: Spreadsheet generation is multi-dimensional. Human preference data captures what users actually value, but different dimensions matter across domains, and some signals surface more clearly than others. Spreadsheet Arena gives us a powerful foundation for evaluation, and a new lens for improving post-training. Start a battle at spreadsheetarena.ai Read the paper at meridian.ai/blog/all/sprea… @srkundurthy @claranahhh @Zachkirshner @calvincbzhang @ManasiSharma_ @jhnling

Feature effects don’t generalize across domains. Finance color coding conventions (e.g., blue inputs, black formulas) aren't significantly impactful on model rankings arena-wide. But zoom into Finance prompts and it's the single strongest predictor of winning. Even then, expert raters disagree with crowd preferences nearly half the time.

Spreadsheets have entered the arena! ⚔️ Announcing Spreadsheet Arena, the first research platform for human preference rankings on LLM-generated spreadsheets. The results? @AnthropicAI Claude Opus is on top, but the gap is tighter than you’d think. w/ @LTIatCMU, @Cornell, and @scale_ai. 🧵