Get a Google Gemini API key
Step-by-step: sign in to Google AI Studio, mint an API key, and drop it into ShellYard Settings → Magellan → Gemini. Free tier covers evaluation.
ShellYard’s Magellan AI requires your own API key when you select the Gemini provider. Keys are minted via Google AI Studio — separate from your Google Cloud Console, even if you already have a GCP project.
Steps
-
Sign in to Google AI Studio. Go to aistudio.google.com and sign in with your Google account. AI Studio is the front door for the Gemini API — separate from the consumer Gemini chat app at gemini.google.com.
-
Create an API key. Open Get API key → Create API key. AI Studio either creates a new Google Cloud project for you or lets you scope the key to an existing project. Name the key something you’ll recognize later — e.g.
shellyard-laptop. Copy it immediately; you can view it again in the AI Studio dashboard later, unlike Anthropic / OpenAI keys. -
Paste it into ShellYard. Open Settings → Magellan → Provider: Gemini → API key and paste. ShellYard fires a small test request to confirm the key works, then stores it (see BYO key for storage details by tier).
Key format
Gemini keys start with AIza followed by ~35 alphanumeric characters.
Picking a model
ShellYard exposes a smart model + fast model pair under the provider settings — Magellan routes per stage. Typical pairings on the Gemini family:
| Tier | Smart model | Fast model |
|---|---|---|
| Best quality | Gemini 2.5 Pro | Gemini 2.5 Flash |
| Balanced (default) | Gemini 2.5 Flash | Gemini 2.5 Flash-Lite |
| Lowest cost | Gemini 2.5 Flash-Lite | Gemini 2.5 Flash-Lite |
The exact model identifiers update over time. See ai.google.dev/gemini-api/docs/models for the current list and context-window sizes.
Pricing
The Gemini API has a free tier with generous daily limits — usually enough for evaluation and light personal use. Paid tier kicks in at higher volumes. Current rate card and free-tier quotas: ai.google.dev/pricing.
Magellan’s five-stage flow attaches scoped context per turn rather than full transcripts, which keeps spend modest. Most infrastructure-engineering workflows land within the free-tier limits.
Rotating
If a key leaks, return to AI Studio → Get API key, revoke the compromised key, mint a new one, and overwrite the value in ShellYard. The revoked key stops working within seconds.
Troubleshooting
- “400 INVALID_ARGUMENT” or “403 PERMISSION_DENIED” — usually means the Gemini API isn’t enabled on the underlying Google Cloud project. AI Studio normally enables it automatically; if not, open Cloud Console → APIs & Services → enable the “Generative Language API” (or “Gemini API”) on the project.
- “429 RESOURCE_EXHAUSTED” — free-tier daily quota hit. Wait for reset, or enable billing on the project for the paid tier.
- Key works in
curlbut not in ShellYard — check Settings → Magellan → Provider is set to Gemini, not OpenAI-compatible.