AI BriefingMercor cuts batch job times from 12 to 2-3 hours using Vercel AI Gateway
·2026.11.03 09:00
Key point
The company routed 18 million requests through Vercel AI Gateway in its first 10 days, enabling over 2,000 internal projects and provisioning expert access in seconds.
Details
- Batch jobs for model evaluation and grading reduced from 12 hours to 2-3 hours by using Vercel AI Gateway for failover routing and increased inference capacity.
- 18 million requests and 63 billion tokens routed through AI Gateway in the first 10 days without hitting rate limits, preventing deadline misses for 24-hour client jobs.
- Over 2,000 internal projects created by hundreds of business team members, using Vercel Passport and a custom product called Audiences to provision access in seconds.
- Per-user keys with expiration dates and spend limits allow engineers to experiment with open models while security teams track usage by team and project.
- Agents are given narrow permissions and scoped model access to limit actions and spending, supporting HR workflows and expert support.
- Mercor connects over 5 million domain experts with AI labs for training and evaluation data generation.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
gpu-unlock Linux module doubles FP/BF16 instruction rate on Nvidia 3000 series GPUs