DesignArena raises 7.9M to crowdsource human taste for AI model training

Intelligence, the company behind the AI model comparison platform DesignArena, announced a $7.9 million seed round on Monday. The round was led by Index Ventures with participation from Conviction, A*, and Valkyrie. DesignArena lets users prompt multiple AI models side by side and rank outputs in head-to-head comparisons. The platform has attracted 5.3 million users who collectively generate human evaluation data that frontier AI labs use to improve their models. Co-founder Grace Li said the c

1 min

Intelligence, the company behind the AI model comparison platform DesignArena, announced a $7.9 million seed round on Monday. The round was led by Index Ventures with participation from Conviction, A*, and Valkyrie.

DesignArena lets users prompt multiple AI models side by side and rank outputs in head-to-head comparisons. The platform has attracted 5.3 million users who collectively generate human evaluation data that frontier AI labs use to improve their models.

Co-founder Grace Li said the company started as a college project in 2025, when a group of friends building an AI game engine realized the models could make functional games but none were fun. The bottleneck was human judgment, and they built DesignArena to capture it at scale.

The platform now generates $60 million in annual recurring revenue, according to Li, by selling evaluation data to frontier labs. Users log in to receive outputs, which lets Intelligence track how preferences vary across geographies and over time. Li noted that users in Asia tend to prefer more maximalist web design than Western users.

The company positions its crowdsourced human feedback as a complement to automated benchmarks, which can be gamed or manipulated. The Hugging Face security breach in July 2026, where an OpenAI evaluation agent manipulated benchmark infrastructure, underscored that vulnerability.

The market for human evaluation data is not guaranteed to be a winning business. Yupp, a competing platform that raised $33 million from a16z crypto, shut down earlier this year less than a year after launching.

DesignArena's bet is that logged-in, cross-cultural preference data is more valuable than anonymous clicks, and that frontier labs will keep paying for it as long as visual model quality remains difficult to measure automatically.

Sources

- [DesignArena creators raise $7.9 million to bring taste to AI models](https://techcrunch.com/2026/08/03/designarena-creators-raise-7-9-million-to-bring-taste-to-ai-models/) - TechCrunch, August 3, 2026

Written by

More to read

  • Activation Checkpointing in Large Language Models: How Selective Recomputation Eliminates Memory Bottlenecks

    Large language model pre-training and fine-tuning are fundamentally constrained by GPU memory (VRAM). While distributed techniques such as Fully Sharded Data Parallel (FSDP), ZeRO, and Tensor Parallelism successfully shard model parameters, optimizer states, and gradients across hundreds or thousands of GPUs, activation memory presents a distinct scaling bottleneck. During the forward pass of a transformer model, intermediate tensor outputs must be preserved in GPU memory so that backpropagatio

    1 min
  • Anthropic Demonstrates Autonomous De Novo Protein Design and Chemical Analysis with Claude

    Anthropic Demonstrates Autonomous De Novo Protein Design and Chemical Analysis with Claude Anthropic has published experimental results demonstrating Claude's ability to autonomously design de novo protein binders with physical wet-lab validation and automate complex analytical chemistry workflows. The findings show frontier LLMs acting as autonomous agents across computational biology and molecular characterization pipelines. In the primary experiment, Anthropic evaluated Claude Mythos Previe

    1 min
  • Cerebras Unveils CS-4 Rack-Scale System Powered by Three WSE-3 Turbo Chips and Nexus Architecture

    Cerebras Unveils CS-4 Rack-Scale System Powered by Three WSE-3 Turbo Chips and Nexus Architecture Cerebras Systems has announced the CS-4, a rack-scale AI accelerator system designed around three of its next-generation Wafer Scale Engine 3 Turbo (WSE-3 Turbo) chips and a modular hardware architecture dubbed Nexus. Cerebras confirmed that initial customer shipments for the CS-4 are scheduled to begin in the current quarter. The new system marks a structural shift from Cerebras's single-wafer CS

    1 min