Vercel Connect pricing row replaced by Token Requests + Triggers
Vercel's pricing page drops the 'Vercel Connect' line item (5K requests/month included, $3 per 10K requests) and replaces it with two separately metered AI Gateway rows: Token Requests ($3.00 per 1,000 requests) and Triggers ($0.95 per 1,000 triggers).
Vercel Connect: 5,000 requests/month included on Hobby; $3 per 10,000 requests overage on Pro; Custom on Enterprise.
Token Requests: 500/month included; $3.00 per 1,000 requests. Triggers: 1,000/month included; $0.95 per 1,000 triggers. Both Custom on Enterprise.
The main pricing page’s AI section previously listed a single “Vercel Connect” row under the AI Gateway feature checklist. As of the August 2026 capture that row is gone, replaced by two distinctly named, distinctly priced metered dimensions — Token Requests and Triggers — with different included allowances and per-unit rates than the product they replaced. Separately in the same capture, Vercel Agent (BETA) added Slack as a third named capability alongside Code Review and Investigations, though that change carried no price impact.
Vercel Sandbox expands from a single region (iad1) to four (iad1, cle1, cdg1, sfo1) and discloses its first regional price differential: Active CPU and Provisioned Memory cost $0.128/hr and $0.0212/GB-hr in Cleveland and Washington D.C., but $0.177/hr and $0.0292-$0.0294/GB-hr (roughly 38% higher) in Paris and San Francisco. Separately, the main pricing page's "Vercel Connect" line item (5K requests/month included, $3 per 10K requests) is replaced by two new metered AI Gateway rows: "Token Requests" (500/month included, $3.00 per 1,000 requests) and "Triggers" (1,000/month included, $0.95 per 1,000 triggers). Vercel Agent (BETA) also adds Slack as a third named capability alongside Code Review and Investigations. Core platform pricing (Pro $20/seat, eight dimensions) is unchanged.