Browse by model family
Every model family and how many glitch tokens are documented for it.
- GPT (OpenAI)15 tokens
- GPT-3.5 / GPT-425 tokens
- DeepSeek3 tokens
- Qwen10 tokens
- Kimi (Moonshot)5 tokens
- Doubao (ByteDance)1 token
- GLM (Zhipu)2 tokens
- Gemini2 tokens
- Llama 35 tokens
- Mistral3 tokens
- Gemma (Google)2 tokens
- MiniMax4 tokens
- Phi-3 (Microsoft)2 tokens
- Command R+ (Cohere)2 tokens
- Llama 26 tokens
- GPT-330 tokens
- GPT-218 tokens
- GPT-J-6B11 tokens
- GPT-NeoX / Pythia1 token
- OLMo 2 (AI2)4 tokens
- Yi (01.AI)1 token
- Falcon 3 (TII)1 token
These tokens have been publicly documented on this model or model family. Inclusion does not mean that every version still shows the same behavior.