Muse Code
Meta's terminal AI coding agent that plans, writes, and validates code across large repositories using persistent background sub-agents.

Data labeling and AI data platform (Meta-invested)
Scale AI remains one of the most capable and largest AI data providers, and for enterprise and government customers it offers scale and quality that few competitors match. The essential 2026 context, though, is Meta's roughly $14.3 billion investment for a large minority stake in 2025, after which several competing AI labs reportedly pulled work over neutrality concerns and Scale trimmed revenue guidance. Buyers should factor in that shift, and note that pricing is custom and enterprise-oriented rather than public.
Scale AI is a data infrastructure company providing labeled training data, RLHF, and evaluation for machine learning and large language models, combining software with a large managed human workforce. In 2025, Meta took a roughly 49% stake for about $14.3 billion, reshaping Scale's client base and strategic position while it continues serving enterprise and government customers.
Scale AI is one of the largest providers of data infrastructure for artificial intelligence. Its core business is producing high-quality labeled training data and human feedback for machine learning models, including annotation for computer vision, autonomous vehicles, and, increasingly, reinforcement learning from human feedback (RLHF) and evaluation for large language models. Scale operates as a vertically integrated platform combining software with a large managed human workforce, and it also offers products for enterprises and government. In June 2025, Meta invested roughly $14.3 billion for a stake reported at around 49%, valuing Scale at over $29 billion. As part of the deal, founder and CEO Alexandr Wang moved to Meta to work on its AI efforts while remaining on Scale's board, and Scale appointed an interim CEO to run the company, which it stated would remain independent. The transaction was a defining event for the data-labeling industry. A notable consequence is that major AI labs that compete with Meta reportedly reduced or ended their reliance on Scale following the deal, prompting Scale to lower revenue expectations and pushing some customers toward alternative providers. Scale continues to serve enterprise and government customers and remains a significant player, but its competitive landscape and neutrality perception shifted materially after Meta's investment. Pricing is enterprise and typically per-task or contract-based rather than public list pricing.
Scale AI is a major AI data infrastructure company providing labeled training data, RLHF, and evaluation, combining software with a large managed workforce. In June 2025, Meta invested about $14.3 billion for a roughly 49% stake at a $29 billion-plus valuation, and its founder joined Meta while Scale stayed independent. Some competing labs reduced use afterward, and Scale trimmed revenue guidance. Pricing is custom and enterprise-oriented.
Scale AI is a US-based data infrastructure company founded in 2016 by Alexandr Wang, providing data labeling, RLHF, and AI data services to enterprises, AI labs, and government.
In June 2025, Meta invested roughly $14.3 billion for a stake reported at about 49%, valuing Scale over $29 billion. Alexandr Wang moved to Meta to work on AI while remaining on Scale's board, and Jason Droege was named interim CEO. Scale stated it would continue to operate independently.
Scale offers large-scale data annotation across text, image, and video, RLHF and human feedback for LLMs, model evaluation and red-teaming, and enterprise and government AI products. It operates a vertically integrated platform pairing software with a managed human workforce accessed via API.
The company monetizes largely per task with contracts often in the six figures, and it serves demanding, high-volume use cases.
Enterprises, AI labs, and government and defense organizations with large-scale data-labeling, RLHF, and evaluation needs.
Machine learning engineers and researchers who consume labeled data, RLHF, and evaluations.
Heads of AI, ML, or data at enterprises and government agencies with large budgets.
Procurement, security, and, post-Meta-deal, competitive-neutrality stakeholders.
A large enterprise or government AI program needing high-volume, high-quality data services and comfortable with Scale's post-Meta strategic position.
Scale AI has raised substantial venture funding over the years. The defining recent event is Meta's June 2025 investment of roughly $14.3 billion for a stake reported around 49%, valuing Scale over $29 billion. Scale remains independent, with its founder now also at Meta.
Meta did not fully acquire Scale. In June 2025, Meta invested roughly $14.3 billion for a stake reported around 49%, valuing Scale over $29 billion. Scale stated it would remain independent, and its founder moved to Meta while staying on Scale's board.
Scale provides labeled training data, RLHF, and evaluation services for machine learning and large language models, combining software with a large managed human workforce, and serves enterprise and government customers.
Reports indicate that some competing AI labs reduced or ended their use of Scale over neutrality concerns after the deal, and Scale lowered revenue expectations, though it continues serving enterprise and government clients.
No. Scale uses custom enterprise pricing, typically per task or by contract, with no self-serve free tier.
It is best for enterprises, AI teams, and government programs with large-scale data-labeling, RLHF, and evaluation needs and the budget for enterprise engagements.
Side-by-side pages for pricing, features, and best-fit use cases.
Meta's terminal AI coding agent that plans, writes, and validates code across large repositories using persistent background sub-agents.
Anthropic's terminal-based coding agent that reads, writes, and refactors across your codebase.
A frontier agentic coding tool that autonomously plans and edits across your codebase from the terminal, editor, web, or phone.
Sourcegraph's AI coding assistant with deep, whole-codebase context.