| AI model | Parameters | Active parameters | Layers | KV heads | Context |
|---|---|---|---|---|---|
| Meta Llama 3.1 8B Instruct | 8 B | – | 32 layers | 8 heads | – |
| Meta Llama 3.1 70B Instruct | 70 B | – | 80 layers | 8 heads | – |
| Meta Llama 3.3 70B Instruct | 70 B | – | 80 layers | 8 heads | – |
| Meta Llama 4 Maverick 17B-128E Instruct | 400 B | 17 B | – | – | 1,000,000 tokens |
| Meta Llama 4 Scout 17B-16E Instruct | 109 B | 17 B | – | – | 10,000,000 tokens |
| Meta Llama-2-7b-chat-hf | 6.74 B | – | – | – | – |
| Meta Llama-2-7b-hf | 6.74 B | – | – | – | – |
| Meta Llama-2-13b-chat-hf | 13 B | – | – | – | – |
| Meta Llama-2-70b-hf | 69 B | – | – | – | – |
| Meta Llama-3.1-8B | 8.03 B | – | – | – | – |
| Meta Llama-3.1-70B | 70.6 B | – | – | – | – |
| Meta Llama-3.1-405B | 406 B | – | – | – | – |
| Meta Llama-3.2-1B | 1.24 B | – | – | – | – |
| Meta Llama-3.2-1B-Instruct | 1.24 B | – | – | – | – |
| Meta Llama-3.2-3B | 3.21 B | – | – | – | – |
| Meta Llama-3.2-3B-Instruct | 3.21 B | – | – | – | – |
| Meta Llama-Guard-3-8B | 8.03 B | – | – | – | – |
| Meta Meta-Llama-3-8B | 8.03 B | – | – | – | – |
| Meta Meta-Llama-3-8B-Instruct | 8.03 B | – | – | – | – |
| Meta Meta-Llama-3-70B | 70.6 B | – | – | – | – |
| Meta Meta-Llama-3-70B-Instruct | 70.6 B | – | – | – | – |
Other makers
- Qwen (Alibaba Cloud)73
- DeepSeek23
- ibm-granite17
- nvidia15
- Microsoft14
- allenai13
- Google13
- zai-org13
- EleutherAI8
- NousResearch8
- Mistral AI7
- LiquidAI6
- HuggingFaceTB5
- moonshotai5
- h2oai4
- MiniMaxAI4
- openbmb4
- poolside4
- bigscience3
- internlm3
- LGAI-EXAONE3
- openai-community3
- tencent3
- TinyLlama3
- XiaomiMiMo3
- 01-ai2
- ai21labs2
- bigcode2
- CohereLabs2
- inclusionAI2
- OpenAI2
- PrimeIntellect2
- tiiuae2
- upstage2
- aisingapore1
- baidu1
- BSC-LT1
- ByteDance-Seed1
- codellama1
- facebook1
- Salesforce1
- stepfun-ai1
- swiss-ai1