Protect
Runs 18+ guardrail types before and after a model call, including PII detection, prompt injection defense, hallucination checks and content moderation, each set to enforce, monitor or log.
What it does
- Guardrails and safety
Sources
- Pricing
-
Pricing page: the free tier lists "Guardrails" at "15 built-in", and the product list names "Guardrails (Protect)".
- Description
-
Company disclosure · 6 Nov 2026
Product page: "18+ guardrail types running pre and post-processing. PII detection, prompt injection defense, hallucination checks, content moderation, secret detection, topic restriction, and custom rules."
- What it does
-
Company disclosure · 6 Nov 2026
Product page: "18+ guardrail types running pre and post-processing. PII detection, prompt injection defense, hallucination checks, content moderation, secret detection, topic restriction, and custom rules."
- Sold within
-
Pricing page: "Usage-based pricing across 6 dimensions." and "Every feature starts free. Only pay when you exceed the generous free tier."
- Name
-
Company disclosure · 6 Nov 2026
Product page: "18+ guardrail types running pre and post-processing. PII detection, prompt injection defense, hallucination checks, content moderation, secret detection, topic restriction, and custom rules."
- URL
-
Company disclosure · 6 Nov 2026
Product page: "18+ guardrail types running pre and post-processing. PII detection, prompt injection defense, hallucination checks, content moderation, secret detection, topic restriction, and custom rules."
- Deployment
-
Company disclosure · 6 Nov 2026
Home page: an "OPEN SOURCE" badge, "Apache 2.0" and a "Self-host it" button.
- Interfaces
-
Company disclosure · 6 Nov 2026
Pricing page: "Full API + SDKs" is listed under "INCLUDED ON EVERY PLAN".
- Intended for
-
Official documentation · 14 Nov 2026
Page: 'Audience Enterprise Scale with confidence'.
Also from Future AGI
-
Command Center API service
Command Center routes LLM requests across providers and applies guardrails, failover, caching, cost controls and request monitoring; also covers PII detection, toxicity blocking and prompt-injection defense at the proxy.
-
Future AGI Platform
Open-source (Apache 2.0) platform for building and testing AI agents that simulates voice and chat agents against personas and generated scenarios, scores them with an evaluation agent that learns from corrections, traces production runs, groups failures into ranked issues with suggested fixes, optimizes prompts and agent configuration and applies guardrails, hosted with a free tier and usage-based billing or self-hosted.
-
Falcon AI AI agent
Agent inside the Future AGI dashboard that you talk to in plain language to analyse evaluation scores and traces, create datasets from production traces and debug why a trace failed, using the platform's own tools plus skills and connected MCP servers.
-
Simulations Developer tool
Runs simulated phone calls and chat conversations against a voice or chat agent using personas and scenarios, recording audio and transcripts and scoring each turn.
Open Future AGI in the directory
Something wrong here? Send a correction — quote this product id: future-agi-protect.