treg

Product Introduction

1. Definition: Treg is an open-source API gateway and unified catalog for AI agent tools. Technically, it functions as an abstraction layer and routing service that standardizes access to a vast ecosystem of third-party APIs.
2. Core Value Proposition: Treg exists to solve the fragmentation and cost barrier in equipping AI agents with specialized capabilities. It provides developers and businesses a single integration point and credential to access thousands of premium endpoints for SEO, social media, data enrichment, and more, billed on a pay-per-call basis instead of requiring multiple expensive subscriptions.

Main Features

1. Unified API Catalog & Key: Treg aggregates 2,630+ endpoints from 47+ specialized providers (e.g., Semrush, Moz, Hunter, SerpApi) into a single, standardized OpenRouter-style interface. How it works: Developers integrate once using a single treg token (trg_live_...). The Treg platform handles all underlying authentication, request translation, and response normalization between its schema and each provider's unique API, eliminating scattered API keys and provider-specific integrations.
2. Evidence-Based Tool Selection: The catalog provides live, measured performance data for each provider's endpoints. How it works: Treg tracks and displays metrics like success rate, speed, and last operational timestamp for comparable tools (e.g., 19 backlink providers). This allows AI agents or developers to make routing decisions based on empirical data—price, reliability, and latency—rather than guesswork. Users can also pin specific vendors for consistent use.
3. Usage-Based Billing & BYOK (Bring Your Own Key): Treg operates on a fractional cent, pay-per-successful-call pricing model, quoting costs upfront. How it works: Treg manages the bulk subscriptions with vendors and passes the volume-based rates to users with a $0.000 markup. If a user already has a subscription (e.g., a Semrush seat), they can inject their own provider key into Treg. The system will prioritize using this key for relevant calls, making those requests unmetered and routing all others through its own pooled infrastructure.

Problems Solved

1. Pain Point: The prohibitive cost and integration complexity of using multiple premium SaaS APIs for AI agents. Agents need diverse capabilities, but purchasing separate seats for Semrush ($139/mo), Moz ($99/mo), Hunter ($34/mo), etc., for sporadic agent use is financially inefficient and technically burdensome.
2. Target Audience: AI Agent Developers building autonomous systems; SaaS Businesses integrating multi-tool capabilities into their products; Growth and SEO Teams requiring programmatic access to data tools without managing dozens of logins; Indie Hackers and Startups needing enterprise-grade data capabilities on a flexible budget.
3. Use Cases: An autonomous SEO agent performing weekly rank tracking, backlink audits, and competitor keyword research using endpoints from SE Ranking, Majestic, and Serpstat in a single workflow. A sales intelligence agent enriching leads with verified emails from Hunter and company data from Crunchbase on-demand. A social media agent publishing content to X, Instagram, and LinkedIn via Postiz endpoints and generating video clips using AI video models.

Unique Advantages

1. Differentiation: Unlike general API aggregation platforms, Treg is purpose-built for the AI agent workflow with upfront pricing, performance-based routing, and a catalog focused on actionable tools (SEO, enrichment, publishing). Compared to manually integrating providers, it reduces development time from months to minutes and converts fixed SaaS costs into variable usage costs.
2. Key Innovation: The open-source, agent-aware routing layer. The entire routing logic and catalog are open-source (AGPL), allowing for auditability and community contributions. Its server-side credential injection ensures API keys never leave Treg's secure environment, and the llms.txt file provides a simple, standardized way for any LLM or agent to discover and interface with the entire tool catalog.

Frequently Asked Questions (FAQ)

1. How does Treg make money if there's no markup on API calls? Treg generates revenue through volume-based pricing agreements with the API providers in its catalog. By aggregating demand, Treg secures wholesale rates and manages the provider subscriptions, earning on the margin between the pooled rate and the pay-per-call price charged to users.
2. Is Treg secure for handling my data and API requests? Yes, Treg is designed with security as a core tenet. Provider credentials are injected server-side and never stored on or exposed to the end-user's machine. The open-source nature of the router also allows for security audits of the routing logic and data handling practices.
3. Can I use Treg if I already pay for a tool like Semrush? Absolutely. Treg supports a "Bring Your Own Key" (BYOK) model. You can add your existing Semrush API key to your Treg account. When your agent makes a call that Semrush can handle, Treg will use your key (unmetered), and will only use its own credits for tools you don't have a subscription for.
4. What happens if an API provider in the Treg catalog changes their interface or goes down? As an open-source project, the community and Treg maintainers update the adapters for each provider. The live performance metrics (success rate, last checked) in the catalog are designed to signal provider health, allowing the routing system or developer to failover to a comparable alternative provider automatically.
5. How do I get my own API or tool listed in the Treg catalog? The Treg catalog is open for submissions. Developers building an API or agent skill can list their tool directly through the Treg website. The process involves contributing to the open-source catalog specifications to ensure standardized integration.