AI API Key Checker

Validate API keys and detect rate limits for 21 AI providers — right inside Obsidian, in one click.

---

✨ Features

• Instant validation — every key is checked against the real provider API, not a regex.
• Rate-limit & balance detection — tells apart a bad key from a good key with no quota, and shows exactly when the limit resets: LIMIT — 17:20 or LIMIT — 07.11 15:00.
• Bulk check — paste 50+ keys at once, get results in seconds.
• Smart sorting — broken keys float to the top so you spot problems first.
• Click to copy — click a masked key to copy it; hover to reveal the full value.
• Stats bar — 12 OK | 3 LIMIT | 2 INVALID (17 total) at a glance.
• Last-check time — see when each provider was last verified.

---

📦 Installation

From Obsidian

Open Settings → Community plugins → Browse, search for AI API Key Checker, and install — or open it directly in the community plugin catalog. Then enable the plugin.

Manual

1. Download main.js, manifest.json, and styles.css from the latest release.
2. Drop them into YourVault/.obsidian/plugins/api-key-checker/:

   YourVault/.obsidian/plugins/api-key-checker/
   ├── main.js
   ├── manifest.json
   └── styles.css
   

3. In Obsidian: Settings → Community plugins → Reload, then enable AI API Key Checker.

Via BRAT

Add this repository in the BRAT plugin to auto-update from releases.

---

🚀 Usage

1. Click the key icon in the left ribbon (or run the command AI API Key Checker: Open panel).
2. Hit Add key, pick a provider, paste one or more keys (one per line).
3. Read the results — grouped by provider, worst status first.

---

🔌 Supported providers

21 providers — click to expand

Provider	Check Method	Notes
OpenAI	/v1/models
Anthropic	/v1/models
Google Gemini	Model list + Gemini probe	Detects daily quota reset
Groq	/v1/models
Grok	Chat completion (grok-3)
FreeModel	Anthropic messages API	Parses reset time from 402
Aerolink	Anthropic messages API
OpenRouter	Chat completion	Shows balance
Kimchi	Anthropic messages API	Detects rate limit via error body
OpenCode Zen	Chat completion (free model)
Xiaomi MiMo	/v1/models	Bearer auth
Amazon Bedrock	Model invoke (nova-lite)	Detects proxy limitations
Z.ai	Chat completion (glm-5)
InferAll	/v1/models
Inception	Chat completion (mercury-2)
Fireworks	Chat completion (deepseek-v4-pro)
Abliteration	Chat completion (abliterated-model)
Anyapi	Chat completion (free model)
Auriko	Chat completion (gpt-3.5-turbo)
v0	/v1/chats
Featherless	Chat completion (Mistral-7B)

---

🎯 Status legend

Status		Meaning
OK	🟢	Key is valid and working
LIMIT	🟡	Key works but is rate-limited or out of credits (shows reset time)
INVALID	🔴	Key rejected — check for typos or expiry
ERROR	🟠	Network issue or the provider is down

---

🔒 Privacy

Your keys never leave your machine except to reach the provider you are validating. There is no telemetry, no analytics, and no third-party server in between. Keys are stored locally in .obsidian/plugins/api-key-checker/data.json — treat that file as a secret and never commit it. The click-to-copy feature writes the selected key to your system clipboard only when you click it; the plugin never reads clipboard content.

---

🛠️ Development

npm install       # install dev dependencies
npm run dev       # watch-mode build (esbuild)
npm run build     # type-check + minified production bundle -> main.js
npm run lint      # ESLint (typescript-eslint + eslint-plugin-obsidianmd)

Source lives in src/ (main.ts = UI/plugin lifecycle, checker.ts = validation engine). Obsidian only loads the bundled main.js, so rebuild after every change. Adding a provider is a five-step change — see CLAUDE.md.

---

📄 License

MIT © VenvStore