# Tessarac

> Tessarac builds infrastructure that pays for itself. Documentation is organized by product: choose the product, then read its index.

## Product documentation

- [AI Cost Optimizer Gateway (ACOG) documentation index](https://tessarac.com/docs/acog/llms.txt): An OpenAI- and Anthropic-compatible gateway in front of the AI providers you already use. It routes each request to the cheapest model that can do the job, caches what can be reused, and enforces your budgets, model access and data loss prevention rules.
- [AI Cost Optimizer Gateway (ACOG) — all documentation in one file](https://tessarac.com/docs/acog/llms-full.txt)
- Security Platform: in beta. https://tessarac.com/security-platform

## Guides

- [How to reduce LLM API costs: a practical checklist](https://tessarac.com/guides/reduce-llm-api-costs): A practical checklist for lowering what you spend on OpenAI, Anthropic, Gemini and other model APIs: measurement, caching, routing, batching and governance.
- [What is an LLM gateway, and when do you need one?](https://tessarac.com/guides/what-is-an-llm-gateway): An LLM gateway sits between your applications and AI model providers. What it does, what to look for, and when one pays for itself.
- [LLM model routing: sending each request to the least costly model that can answer it](https://tessarac.com/guides/llm-model-routing): Most requests do not need your most capable model. How difficulty-based model routing works, where it goes wrong, and how to route without hurting quality.
- [Prompt caching explained: Anthropic, OpenAI and Gemini compared](https://tessarac.com/guides/prompt-caching): How prompt caching works on each major provider, what has to stay identical for a cache hit, and how to structure prompts so repeated context costs less.
- [LLM batch APIs: the discount for work nobody is waiting on](https://tessarac.com/guides/llm-batch-apis): Anthropic, OpenAI and Google all price batch requests below real-time ones. What the batch tiers trade away, which workloads fit, and how to use them safely.
- [Semantic caching for LLMs: answering repeat questions without a model call](https://tessarac.com/guides/semantic-caching-for-llms): Response caching answers identical requests; semantic caching also answers questions that mean the same thing. How it works, where it is safe, and the traps.
- [AI data loss prevention: keeping sensitive data out of LLM prompts](https://tessarac.com/guides/keep-sensitive-data-out-of-llm-prompts): Card numbers, credentials, customer records and source code end up in prompts. How data loss prevention for LLM traffic works and how to roll it out.
- [AI costs out of control? Budgets, attribution and visibility for LLM spend](https://tessarac.com/guides/control-ai-spend-budgets-and-attribution): When the AI line on the invoice grows faster than anyone can explain, start with attribution and budgets. How to see who spends what, and cap it safely.
- [Consolidating identity, secrets and privileged access onto one platform](https://tessarac.com/guides/consolidating-identity-secrets-and-access): Identity, secrets management and privileged access are usually three vendors with three audit logs. What consolidating them changes, and what to check first.

## Site

- [Documentation home](https://tessarac.com/docs): choose a product
- [Guides](https://tessarac.com/guides): articles on LLM cost and AI security