AI Provider Setup · BYOK
One key, every frontier model — the fastest way to try them all.
OpenRouter proxies dozens of model providers behind a single API key, so you can switch between Claude, GPT, Gemini, Grok, DeepSeek, and Llama without signing up for each one. It is the default we recommend for most Admaxxer users — pay-as-you-go, no monthly minimums, and you can change models per-chat without touching settings.
OPENROUTER_API_KEYSign up at openrouter.ai with Google, GitHub, or email. Free to create the account — you only pay for tokens you actually use.
Go to openrouter.ai/credits and top up with a card. $10 is enough to test every model on Admaxxer for a few weeks of normal use.
Navigate to openrouter.ai/keys or click your avatar then Keys. OpenRouter lets you create multiple keys so you can scope one per app.
Click Create Key, name it Admaxxer, optionally set a credit limit (e.g. $25), and hit Create. The limit protects you if the key ever leaks.
The full key is shown exactly once. It begins with sk-or-. Paste it somewhere safe before closing the dialog.
Go to Settings then AI Providers, find the OpenRouter row, paste the key, and click Connect. Admaxxer runs a test call and stores the key encrypted.
We recommend anthropic/claude-opus-4-7 for best reasoning plus tool use. You can always override the model per chat.
Re-check the key starts with sk-or-, that you have credits loaded, and that the key has not been revoked on openrouter.ai/keys.
Once you've pasted your key into /settings/ai-providers, hit Test. Admaxxer makes a single no-cost call against the provider's /v1/models (or equivalent) endpoint and reports back:
These are the pinned model IDs Admaxxer tests against. The model picker in chat will show these plus any live-catalog entries the provider returns.
| Model ID | Display name | Context | Tools | Best for |
|---|---|---|---|---|
anthropic/claude-opus-4-7 |
Claude Opus (via OpenRouter) (recommended) | 200K | Yes | Best all-round — deep reasoning, tool use, and 200K context. Our default pick. |
anthropic/claude-sonnet-4-6 |
Claude Sonnet (via OpenRouter) | 200K | Yes | Faster and cheaper than Opus. Good for high-volume agent runs. |
openai/gpt-5 |
GPT-5 (via OpenRouter) | 400K | Yes | Largest context on OpenAI side. Useful for long-document analysis. |
google/gemini-2.5-pro |
Gemini 2.5 Pro (via OpenRouter) | 1000K | Yes | 1M-token context for very large dumps — ad creative libraries, transcripts, logs. |
xai/grok-4 |
Grok 4 (via OpenRouter) | 256K | Yes | Real-time X/Twitter access. Useful when questions depend on today's chatter. |
deepseek/deepseek-r2 |
DeepSeek R2 (via OpenRouter) | 128K | Yes | Cheapest reasoning tokens available. Good for bulk analytical workloads. |
When it happens: On first Connect, or later if the key is revoked.
Fix: Regenerate the key at openrouter.ai/keys and paste the new one. Keys cannot be recovered — rotate, don't retry.
When it happens: After your credits drop to zero.
Fix: Top up at openrouter.ai/credits. OpenRouter does not auto-recharge by default.
When it happens: Under heavy concurrent load on a specific model.
Fix: OpenRouter rate-limits per-model. Slow down, or switch to a different model for the current chat.
When it happens: If the selected model id is mistyped or has been deprecated upstream.
Fix: Pick a model from Admaxxer's list or browse openrouter.ai/models for current ids.
When it happens: If the upstream provider (Anthropic, OpenAI, etc.) is down.
Fix: Check openrouter.ai/status. Switch to another provider's model in the meantime.