How-To Series · 8: Provider Plumbing

Google Vertex A I Setup

Gemini models on G C P with enterprise rate limits. OAuth 2 tokens, auto-refreshed. No static key.

After this videoYou can wire Hermes to Google Vertex AI and let it handle OAuth 2 token refresh automatically.

Hermes Agent supports Gemini models on Google Cloud Vertex AI through an OpenAI-compatible endpoint. Unlike the A I Studio provider, Vertex uses OAuth 2 short-lived tokens, not a static API key, and draws on G C P billing. This guide covers the two credential paths (service account J S O N or Application Default Credentials), the hermes model picker, where each setting lives, and the troubleshooting fixes for common errors.

About these resources. This episode explains how to set up Hermes with Google Vertex AI. The Sources block names the Hermes docs page that backs every claim. For the A I Studio alternative, see the related guide.

New words here · Plain English

one sentence each · full glossary
Vertex AIGoogle Cloud platform for running AI models with enterprise-level rate limits and OAuth security.
OAuthA secure login flow where one app gets permission to use another without seeing your password.

Sources · What this video distills

3 docs pages · every command below traces to one of them
Primary, the full Vertex setup guide
Google Vertex AI
Read ↗
The static-key Gemini alternative referenced for comparison
Google Gemini (A I Studio)
Read ↗
Configuration split between dot env and config dot yaml
Configuration
Read ↗

Commands shown · Copy and paste

each shows the source doc it came from
Set service account credentials (Option A)from source ↗
echo "VERTEX_CREDENTIALS_PATH=/path/key.json" >> ~/.hermes/.env
Application Default Credentials (Option B)from source ↗
gcloud auth application-default login
Select Vertex as your providerfrom source ↗
hermes model
Diagnose Vertex configurationfrom source ↗
hermes doctor

Going deeper · Related Hermes docs

further reading · not sources of facts shown above