Corrath

Corrath is an AI gateway that acts as a transparent, high-performance proxy between applications and large language model (LLM) providers. It provides deterministic security filtering, spending controls, and automated failover to protect LLM-powered applications from adversarial threats, budget overruns, and provider outages. The platform integrates via a unified API endpoint, enabling developers to route requests to multiple model backends (including OpenAI, Anthropic, Google, Mistral, and DeepSeek) without changing application code. Key features include real-time prompt injection neutralization, granular per-key spending limits, and a sub-5ms latency circuit breaker that dynamically reroutes traffic during provider downtimes. Corrath also offers cost forecasting, HMAC-SHA256 webhook alerts for threat and cost anomalies, and a dashboard for monitoring request distributions, token spend, and threat categories. Designed for enterprise production environments, Corrath claims 99.99% uptime SLA and processes over 500,000 protected requests and 10 million tokens. It supports multiple project spaces, vaulted API keys, and role-based access controls (RBAC) in higher tiers. An enterprise plan offers on-premise deployment and dedicated support.

Category: AI API
Pricing: Unknown

Problem solved

Direct integration with LLM APIs exposes applications to prompt injection attacks, budget-busting runaway completions, provider outages that crash services, and compliance risks from leaked PII or credentials. Corrath eliminates these risks by sitting as a secure, low-latency proxy that inspects every request and enforces security, cost, and availability policies.

Use cases

Protecting chatbots and customer-facing AI assistants from prompt injection and jailbreaks
Enforcing spending limits across development teams to prevent accidental cloud cost spikes
Maintaining application uptime by automatically switching LLM providers during outages
Auditing all LLM requests for PII, security threats, and compliance violations
Monitoring real-time token usage and latency across multiple AI services
Integrating a secure gateway for internal agentic workflows that call multiple models

Audience

Developers
ML Engineers
Data Scientists
Enterprise

Pros

Sub-5ms latency overhead ensures minimal impact on application performance
Turnkey protection against prompt injection and adversarial attacks
Built-in spending controls and cost forecasting prevent budget overruns
Automatic failover across multiple LLM providers for high availability
Comprehensive monitoring and webhook alerts for security and cost anomalies

Cons

Free tier limited to 500 requests per month, unsuitable for production
Pricing details for paid plans are not publicly listed
Limited to a fixed set of supported model providers
Adds an external dependency in the request path, introducing a potential single point of failure
No open-source version or self-hosted option outside of Enterprise plan

Official website: https://www.corrath.io/

Canonical LambdaBase page: https://www.lambdabase.com/tools/corrath

← Directory

Corrath

Unknown

• Updated Jul 20, 2026, 03:06 AM

Corrath is an AI gateway that acts as a transparent, high-performance proxy between applications and large language model (LLM) providers.

Use Cases ▼

Protecting chatbots and customer-facing AI assistants from prompt injection and jailbreaks

Enforcing spending limits across development teams to prevent accidental cloud cost spikes

Maintaining application uptime by automatically switching LLM providers during outages

Auditing all LLM requests for PII, security threats, and compliance violations

Monitoring real-time token usage and latency across multiple AI services

Integrating a secure gateway for internal agentic workflows that call multiple models

What Problem It Solves ▼

Key Features ▼

✓

Sub-5ms latency overhead ensures minimal impact on application performance

✓

Turnkey protection against prompt injection and adversarial attacks

✓

Built-in spending controls and cost forecasting prevent budget overruns

✓

Automatic failover across multiple LLM providers for high availability

✓

Comprehensive monitoring and webhook alerts for security and cost anomalies

Overview

Full Description ▼

Key features include real-time prompt injection neutralization, granular per-key spending limits, and a sub-5ms latency circuit breaker that dynamically reroutes traffic during provider downtimes. Corrath also offers cost forecasting, HMAC-SHA256 webhook alerts for threat and cost anomalies, and a dashboard for monitoring request distributions, token spend, and threat categories.

Designed for enterprise production environments, Corrath claims 99.99% uptime SLA and processes over 500,000 protected requests and 10 million tokens. It supports multiple project spaces, vaulted API keys, and role-based access controls (RBAC) in higher tiers. An enterprise plan offers on-premise deployment and dedicated support.