Releases

Scale completed SWE Atlas by adding Test Writing and Refactoring alongside Codebase QnA, producing a 284-task software-engineering evaluation suite.

Scale Labs updated MultiChallenge with a stronger judge model, refined tasks, and refreshed results to improve agreement with expert ratings.

Scale expanded SEAL into Scale Labs, a broader research organization focused on evaluation, agents, multimodal systems, post-training, deployments, and oversight.

Scale launched high-fidelity simulated environments for training and testing agents on tool use, computer use, coding, and professional workflows.

Scale expanded the SEAL Leaderboards into agent evaluation, beginning with SWE-Bench Pro and MCP Atlas.

Scale formalized a data-engine offering for robotics and embodied AI, covering multimodal collection, annotation, and real-world model improvement.

Scale added richer model comparison, failure analysis, cross-sectional reporting, and automated error-discovery capabilities to its frontier evaluation framework.

Scale made its model-evaluation platform generally available for model developers, enterprises, and public-sector organizations.

Scale launched expert-evaluated leaderboards for coding, instruction following, mathematics, and multilinguality using private, contamination-resistant prompt sets.

Scale expanded its generative-AI platform and services into custom agentic applications connected to enterprise data, controls, and workflows.

Scale launched its Safety, Evaluations, and Alignment Lab to build standardized frontier-model benchmarks, evaluation products, and red-team methods.

Scale introduced an expert-driven testing and evaluation platform for measuring frontier-model capability, safety, bias, and robustness.

Scale launched EGP as a model-agnostic enterprise platform for deploying secure generative-AI applications with proprietary data, fine-tuning, evaluation, and red teaming.

Scale launched Donovan, a secure decision-support platform using large language models, retrieval, and mission data for government and defense users.

Scale launched Outlier to connect subject-matter experts with generative-AI training and evaluation projects.

Scale expanded its data engine beyond perception data into expert demonstrations, preference feedback, red teaming, and reinforcement learning from human feedback for language models.

Scale moved Rapid into general availability for fast labeling experiments without data minimums.

Scale released the core Nucleus workspace for dataset visualization, search, curation, collaboration, and model-error analysis.

Scale launched Nucleus as a dataset-management and model-debugging workspace for visualizing, curating, searching, and collaborating on machine-learning data.

Scale introduced a self-serve, pay-as-you-go product for receiving production-quality labels and instruction feedback in as little as one hour.

Scale launched a mapping product for computer-vision and autonomous path-planning workflows.

Scale introduced a managed workspace for configuring annotation projects, instructions, quality controls, and review workflows.

Scale expanded its annotation platform around the image, video, lidar, and sensor-fusion requirements of autonomous-vehicle developers.

Scale began building distributed contributor platforms and training operations to support large-scale annotation, expert-data, and evaluation programs.

Scale launched Remotasks to train and coordinate distributed contributors performing data annotation and review tasks.

Scale launched the data infrastructure that became its core Data Engine, combining APIs, machine assistance, and human review to create high-quality training data.