标题: Wirestock完成2300万美元A轮融资服务6家大模型厂商提供多模态数据
摘要:
Wirestock宣布完成2300万美元A轮融资,用于扩展其数据供应业务。本轮由Nava Ventures领投,SBVP、Formula VC和I2BF Ventures参投。该公司自2023年转型为AI数据提供商,现向6家头部基础模型公司供应多模态数据。
Wirestock早期主要帮助摄影师向Shutterstock等图库平台分发和销售作品,转型后开始提供图像、视频、设计素材、游戏内容和3D内容数据集。公司表示,平台已有超过700000名艺术家和设计师注册,参与数据采集相关任务。其年化收入运行率为4000万美元,累计向贡献者支付1500万美元。
公司称已允许创作者选择退出数据供应业务,并强调转型过程保持透明。随着AI实验室对定制内容和标注数据需求增长,创意平台正将存量内容资产转化为训练数据供应能力,这反映出多模态数据采购正在成为生成式AI产业链的重要环节。
创意平台加速转向数据供应
多模态数据需求持续上升
定制标注服务成为新增长点
Wirestock Raises 23M Series A Multimodal Data Platform 700000 Creators
Wirestock has raised $23 million in Series A funding to expand its AI data supply business. The company, which previously focused on distributing stock content for photographers, pivoted in 2023 to supplying multimodal datasets to AI labs. Wirestock said its platform now includes more than 700,000 artists and designers, serves six large foundation model developers, has a $40 million annual run-rate revenue, and has paid $15 million to contributors.
The company provides multimodal data, including images, videos, design assets, gaming content, and 3D content, for foundation model training. According to CEO Mikayel Khachatryan, the business initially licensed existing content libraries, then shifted toward custom data requests from AI labs, creating a task-based workflow similar to freelancer marketplaces. Wirestock said it informed creators about the transition and allowed them to opt out of AI data licensing, although it did not disclose the exact share of contributors who joined the new model. As part of the pivot, the company retrained teams to perform annotation and data labeling to make datasets usable for machine learning pipelines.
The funding round was led by Nava Ventures, with participation from SBVP, Formula VC, and I2BF Ventures. Wirestock’s growth reflects a broader market trend in which content platforms monetize proprietary creator networks and structured media libraries as training data infrastructure for generative AI and other foundation models.
Key Takeaways:
Wirestock shifted from stock distribution to AI data infrastructure in 2023
The company combines creator sourcing with annotation for custom training datasets
Marketplace platforms are increasingly monetizing proprietary media as model training data
Source: Original Article
查看原文 →
View Original →