---
title: "Overview"
description: "Request-level capabilities of the chat completions endpoint — caching, vision, tools, metadata, compression, compaction, and skill prompts."
---

> Documentation Index
> Fetch the complete documentation index at: https://truememory-docs.mvstg.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

Every feature below is a capability of the [Unified API](/ai-gateway/chat-completions/) — you opt in per request with a parameter or header, no configuration step required.

| Feature | What it does | Opt in with |
|---------|--------------|-------------|
| [Prompt caching](/ai-gateway/features/prompt-caching/) | Reuse a stable prompt prefix across requests at up to 10× lower cost (Anthropic routes) | `cache_control` on a message |
| [Vision (image input)](/ai-gateway/features/vision/) | Send images to vision-capable models alongside text | content parts in `messages` |
| [Tool calling](/ai-gateway/features/tool-calling/) | Let the model call functions your app defines and executes | `tools` |
| [Custom metadata](/ai-gateway/features/custom-metadata/) | Attach key-value pairs to a request, persisted with the thread and echoed back | `metadata` |
| [Message compression](/ai-gateway/features/message-compression/) | Summarise older turns in long threads to reduce token usage | `"compression": "true"` in `metadata` |
| [Thread compaction](/ai-gateway/features/thread-compaction/) | Fold older turns into a rolling summary checkpoint so long-thread cost stays flat | `context_management` |
| [Skill instructions](/ai-gateway/features/skill-instructions/) | Inject caller-composed skill prompts without replacing the agent's identity | `skill_instructions` |

Features compose freely — a single request can cache its system prompt, send an image, define tools, and carry metadata at the same time.

Source: https://truememory-docs.mvstg.com/ai-gateway/features/overview/index.mdx