VercelVercel added models to AI Gateway and Token Pass, increased build disk space and sped up Sandbox snapshots. The updates cover coding agents, AI apps and deployment infrastructure.
Claude Fable 5.2 Joins AI Gateway
Anthropic’s Claude Fable 5.2 is now available through AI Gateway for open-ended coding, reasoning and vision work. It supports effort levels from low to max, defaults to high, and requires 30-day retention for prompts and completions, which are not used for training. Zero Data Retention is not supported.
Build Disk Space Doubles to 64 GB
Vercel has increased build disk space from 32 GB to 64 GB on Basic and Standard machines, matching Enhanced and Turbo. The increase applies automatically across every plan and build machine type at no extra charge.
Muse Ember Arrives on AI Gateway
Meta’s Muse Ember is now available on AI Gateway for high-volume agent work such as coding sub-tasks, tool calls and document analysis. It accepts text, image and video input, has a 1M-token context window, and supports adjustable reasoning from minimal to max.
GLM 5.3 Flash Enters Token Pass
Z.ai’s multimodal coding model GLM 5.3 Flash has joined Vercel’s Token Pass, a $30 monthly subscription for Pro teams that includes $80 in monthly credit. It has a 1M-token context window, works with coding agents and custom AI SDK apps, and uses Zero Data Retention by default, following Vercel’s move to apply Token Pass credit automatically in AI Gateway.
Sandbox Snapshots Get Up to 10x Faster
Vercel Sandbox snapshot creation is now up to 10x faster, with the median time falling from 4.4 seconds to under 0.4 seconds. Restore time at p50 has dropped from 500ms to 250ms, while persistent sandboxes adopt the new Vercel Hive Snapshot format automatically.
Sources
5 checkedHow we cover tool news: Create With's tool desk drafts these reports with AI from the sources listed above and checks them against those sources before publishing.
Jotform
