What Token NomNom is
A game that turns leftover AI API tokens into points. Connect a provider, burn tokens, earn in-app coins, unlock achievements and compete on a live leaderboard.
Every other tool in the AI developer ecosystem exists to help you use fewer tokens. Token NomNom — created by Mothilal-M — turns the entire premise on its head. It takes the metric developers try to minimize and makes it the high score.
How the Game Works
- Connect Any AI Provider: Bring your own API key for OpenAI, Anthropic Claude, Google Gemini, Groq, Mistral, DeepSeek, or any local endpoint (Ollama/vLLM).
- Stream & Incinerate: Send prompts through the animated combustion chamber with live token pyrometers and rising ember particle effects.
- Earn In-App Coins: Validated token counts are awarded as coins into an append-only transaction ledger with combo multipliers.
- Dominate the Leaderboard: Compete across Daily, Weekly, Monthly, and All-Time leaderboards, challenge rivals, and defeat community World Bosses.
Supported AI Providers Comparison
| Provider | Key Models | Special Tracking | Pricing Tier |
|---|---|---|---|
| OpenAI | GPT-4o, GPT-4o-mini, o1, o3-mini | Input, Output, Cached, Reasoning | $0.15 - $5.00 / 1M input |
| Anthropic | Claude 3.5 Sonnet, Claude 3.5 Haiku | Input, Output, Cache Read/Write | $0.25 - $3.00 / 1M input |
| Google Gemini | Gemini 1.5 Flash, Gemini 1.5 Pro | Prompt, Candidates, Cached | $0.075 - $1.25 / 1M input |
| Groq | Llama 3.3 70B, Mixtral 8x7B | Sub-second live streaming | $0.05 - $0.59 / 1M input |
| DeepSeek | DeepSeek V3, DeepSeek R1 | Reasoning tokens, low-cost burn | $0.14 - $0.55 / 1M input |
| Local / Ollama | Llama 3, Mistral, Qwen, vLLM | Direct local GPU tracking | $0.00 (100% Free) |
Security & Spend Guardrails
The coins are game points, but your API tokens cost real money from your provider. Token NomNom includes strict hardware-grade safety:
- AES-256-GCM Encryption: API keys are encrypted at rest and decrypted in RAM only when making outbound requests.
- Zero Client Storage: Keys are never saved to browser storage and never accessible to client scripts.
- Custom Spend Ceilings: Configure hard limits on tokens per request, daily request quotas, and estimated daily dollar spend caps.
- Preflight Cost Calculator: Every prompt calculates an estimated token count and projected cost before execution.