# Ascend Gateway V5

> One endpoint. 34 MCP tools. 70+ SaaS APIs. Sub-10ms gateway overhead.

## Docs

- [Architecture](https://docs.ascendgtm.net/docs/explanation/architecture.md): How the V5 Gateway works under the hood.
- [Security Model](https://docs.ascendgtm.net/docs/explanation/security.md): Defense-in-depth security for the V5 Gateway.
- [Authentication](https://docs.ascendgtm.net/docs/guides/authentication.md): Bearer tokens and OAuth 2.1 for the V5 Gateway.
- [Connect to Amazon Q](https://docs.ascendgtm.net/docs/guides/connect-to-amazon-q.md): Use V5 Gateway tools in Amazon Q Developer and Q Business
- [Connect to ChatGPT](https://docs.ascendgtm.net/docs/guides/connect-to-chatgpt.md): Use Ascend Gateway V5 as an MCP server in ChatGPT.
- [Connect to Claude](https://docs.ascendgtm.net/docs/guides/connect-to-claude.md): Use Ascend Gateway V5 as an MCP server in Claude Desktop, Claude Code, or claude.ai.
- [Connect to Codex CLI](https://docs.ascendgtm.net/docs/guides/connect-to-codex.md): Use V5 Gateway tools in OpenAI Codex CLI agentic sessions
- [Connect to Cursor](https://docs.ascendgtm.net/docs/guides/connect-to-cursor.md): Use Ascend Gateway V5 as an MCP server in Cursor IDE.
- [Connect to Gemini](https://docs.ascendgtm.net/docs/guides/connect-to-gemini.md): Use Ascend Gateway V5 with Google Gemini CLI or SDK.
- [Connect to LM Studio](https://docs.ascendgtm.net/docs/guides/connect-to-lm-studio.md): Use V5 MCP tools in LM Studio's agentic chat.
- [Connect to Mistral](https://docs.ascendgtm.net/docs/guides/connect-to-mistral.md): Use V5 MCP tools in Mistral Le Chat and the Mistral API.
- [Connect to n8n](https://docs.ascendgtm.net/docs/guides/connect-to-n8n.md): Call V5 Gateway tools from n8n workflows via REST API
- [Connect to Open WebUI](https://docs.ascendgtm.net/docs/guides/connect-to-open-webui.md): Use V5 MCP tools in Open WebUI with any backend model.
- [Connect to Perplexity](https://docs.ascendgtm.net/docs/guides/connect-to-perplexity.md): Use Ascend Gateway V5 as an MCP server in Perplexity.
- [Connect to Slack](https://docs.ascendgtm.net/docs/guides/connect-to-slack.md): Query Ascend Gateway V5 directly from Slack via the GTM bot.
- [Connect to VS Code / GitHub Copilot](https://docs.ascendgtm.net/docs/guides/connect-to-vscode.md): Use V5 Gateway tools in VS Code with GitHub Copilot Chat
- [Connect to Windsurf](https://docs.ascendgtm.net/docs/guides/connect-to-windsurf.md): Add V5 Gateway as an MCP server in Codeium Windsurf
- [Connect to Zed](https://docs.ascendgtm.net/docs/guides/connect-to-zed.md): Add V5 Gateway as an MCP server in the Zed editor
- [Your First API Call](https://docs.ascendgtm.net/docs/guides/first-api-call.md): Make a real API call and understand the response format.
- [Async Jobs](https://docs.ascendgtm.net/docs/guides/hermes/async-jobs.md): Kick off long-running agent tasks in the background with optional webhooks
- [Chat Sessions](https://docs.ascendgtm.net/docs/guides/hermes/chat-sessions.md): Start and continue multi-turn conversations with Hermes agents
- [Orchestrator Pattern](https://docs.ascendgtm.net/docs/guides/hermes/orchestrator.md): Route requests across specialized agents using a classifier front-end
- [Hermes — Autonomous Agent Layer](https://docs.ascendgtm.net/docs/guides/hermes/overview.md): Run persistent AI agents over your GTM data without managing infrastructure
- [Scheduled Jobs](https://docs.ascendgtm.net/docs/guides/hermes/scheduled-jobs.md): Run agent workflows automatically on daily or weekly cadences
- [Multi-Account Support](https://docs.ascendgtm.net/docs/guides/multi-account.md): Manage multiple accounts per provider per tenant.
- [Quick Start](https://docs.ascendgtm.net/docs/guides/quickstart.md): Get your first API call working in 5 minutes.
- [Admin API Reference](https://docs.ascendgtm.net/docs/reference/admin-api.md): Protected endpoints for tenant and token management.
- [llms.txt + MCP from Docs](https://docs.ascendgtm.net/docs/reference/llms-txt.md): Machine-readable docs and the auto-generated MCP server from Mintlify
- [MCP Tools Reference](https://docs.ascendgtm.net/docs/reference/mcp-tools.md): All 34 tools available via the MCP endpoint with parameters and examples.
- [Providers](https://docs.ascendgtm.net/docs/reference/providers.md): 70+ API providers available through the gateway.
- [Quality Harness](https://docs.ascendgtm.net/docs/reference/quality-harness.md): How V5 monitors and grades every tool call for accuracy and safety
- [Recipes](https://docs.ascendgtm.net/docs/reference/recipes.md): Pre-built multi-step workflows composing multiple API calls.
- [Response Trimming](https://docs.ascendgtm.net/docs/reference/response-trimming.md): Reduce token usage by 40-80% with fields and format parameters.
- [REST API Reference](https://docs.ascendgtm.net/docs/reference/rest-api.md): Traditional HTTP API for non-MCP consumers.
- [Ascend Gateway V5](https://docs.ascendgtm.net/index.md): One endpoint. 34 MCP tools. 70+ SaaS APIs. Sub-10ms gateway overhead.

## OpenAPI Specs

- [openapi](https://docs.ascendgtm.net/api-reference/openapi.json)