Arena's $100M revenue challenge: can crowdsourced quality beat Scale AI?
arena hit $100m in annualized revenue from its ai model evaluation service, positioning it as a major competitor to established data labeling giants like scale ai.
arena, originally a research project at uc berkeley, reached $100 million in annualized revenue by selling deep-dive performance analytics to model labs and enterprises who previously relied on its free consumer-facing leaderboard. this milestone, achieved only eight months after launching its commercial service, suggests a massive shift in the ai market where developers prioritize high-quality, crowdsourced feedback over traditional labeling vendors like scale ai and mercor. with competitors like handshake and mercor also seeing revenue surges, the post-training refinement market is booming, though arena faces the unique challenge of converting consumption-based revenue into a stable recurring stream.