Safeguard today removed every limit on MCP tool calls. Teams that connect Safeguard to Claude, ChatGPT, Cursor or their own agents can now call its tools as often as their work needs, on every plan, Free included. There is no monthly cap, no per-minute peak and no charge for the calls.
AI assistants are becoming the place where engineers ask security questions. Is this package safe? Which of these CVEs can actually be reached? What changed in this dependency? A metered tool call turns each of those questions into a budget decision, and the answer is usually to ask less often. Safeguard thinks that is the wrong trade.
"An agent that has to ration its security checks will skip them," said Hritik Kumar Sharma, founder of Safeguard. "We want the check to be the cheapest thing in the loop."
Alongside the change, the Safeguard API moves to pay as you go, with no monthly API plan to buy. The first 10,000 deterministic calls each month are included, then scans and lookups cost $0.10 per 1,000 calls. Model usage carries no markup: customers pay the provider's list price.
What's included
- Unlimited MCP tool calls on Free, Starter, Pro, Plus and Max plans.
- Pay-as-you-go API pricing with 10,000 deterministic calls included each month.
- No markup on model usage.
Full details are on the pricing page at safeguard.sh/pricing.
About Safeguard
Safeguard is the AI-native software supply chain security platform. It finds risk across code, dependencies, containers and AI, and fixes it autonomously with tested pull requests. It can be deployed as SaaS, in a private cloud, on-prem, or fully air-gapped. Founded in 2024, Safeguard is headquartered in Dublin, California. Learn more at safeguard.sh.
