Cloudflare AI Gateway Adds Real-Time Spend Limits, Identity-Driven Budgets

Cloudflare

Jun 5, 2026 · Updated Jun 13, 2026

Cloudflare's AI Gateway now offers real-time spend limits to control token costs across various AI providers. This update, combined with identity-driven budgets via Cloudflare Access, helps companies manage AI expenses and attribute usage to specific users or teams. It addresses the challenge of runaway bills and lack of visibility in enterprise AI deployments.

Cloudflare AI Gateway now includes real-time spend limits, allowing organizations to set dollar-based budgets for AI usage across multiple providers and models. Limits can be scoped by model, provider, or custom attributes (user, team, application), with daily, weekly, or monthly windows. When a budget is met, the gateway can block requests or reroute them to a fallback model.

This update addresses unmanaged AI costs and lack of visibility, often leading to overspending on frontier models for routine tasks. By providing granular cost control and attribution, AI Gateway helps companies track ROI and manage AI expenses like other business line items. It builds on its existing capabilities for unified billing and logging across multiple providers.

Spend limits are in open beta now, set from the dashboard or the API. A closed beta adds identity-driven budgets through Cloudflare Access, so a limit can attach to a specific user, team, or even an autonomous agent — with the cost broken down the same way.

View the full update on blog.cloudflare.com

Cloudflare

@CloudflareJun 5

AI Gateway now features real-time spend limits to prevent runaway token bills across multiple AI providers. By integrating with Cloudflare Access, companies can use identity-driven budgets and policies. https://t.co/tsW1Q5Vhso

21184

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from Cloudflare →

Keep reading

Cloudflare adds xAI Grok models to AI Gateway with unified billing

Cloudflare has integrated xAI's full Grok model suite into its AI Gateway platform. This move allows developers to deploy frontier reasoning and generative media models through a single control plane, eliminating the need for separate API management and fragmented billing across providers.

Cloudflare Integrates GPT-5.5 to Power Persistent Autonomous Agents

CloudflareApr 24

Cloudflare Integrates GPT-5.5 to Power Persistent Autonomous Agents

Cloudflare added OpenAI's GPT-5.5 to its AI Gateway, featuring a 1M token context window and 2x cost efficiency over competing frontier coding models. The model is optimized for agentic loops, enabling systems to plan, use tools, and self-verify their work until a task is complete.

Vercel AI Gateway Automates Provider Selection Based on Cost and Performance

VercelMay 18

Vercel AI Gateway Automates Provider Selection Based on Cost and Performance

Vercel added a programmatic sort feature to its AI Gateway that ranks model providers by cost, latency, or throughput at the request level. This update moves the logic of AI economics into the infrastructure layer, allowing applications to automatically route to the cheapest or fastest available provider without code changes.

Vercel AI Gateway Recovers Over 1 Trillion Tokens Monthly with Built-in Redundancy

Guillermo RauchJun 8

Vercel AI Gateway Recovers Over 1 Trillion Tokens Monthly with Built-in Redundancy

Vercel announced its AI Gateway recovers over 1 trillion tokens monthly by providing redundancy and failover mechanisms. The service offers zero markup over model providers, along with zero-data retention enforcement, observability, usage APIs, and caps. This positions the gateway as a resilient and cost-effective solution for deploying AI applications.