New AI Models and cheaper pricing


New Models (Website and API):
• GPT-4O Mini
• Llama 3.1 8B
• Llama 3.1 70B
• Llama 3.1 405B
• Gemini 1.5 Flash
• Qwen 2 72B
• Deepseek LLM 67B
• Gemma 2 9B
• Gemma 2 27B

Deprecated Models (Website Only, Still on API):
• Llama 2 (7B, 13B, 70B)
• Codellama (7B, 13B, 34B, 70B)

Now Free:
• Claude 3 Haiku (Previously Paid)

LLM Pricing (USD per 1M tokens):
4-9B: Input 0.15, Output 0.3
9-21B: Input 0.3, Output 0.45
21-41B: Input 0.8, Output 1.2
41-80B: Input 1, Output 1.5
80-110B: Input 1.5, Output 2.5
110-320B: Input 3, Output 5
320-405B: Input 5, Output 8
Website: Free for 4-9B models

MoE Model Pricing:
8x7B: Input 0.3, Output 0.45
8x22B: Input 1, Output 1.5

Note: Closed model pricing remains unchanged.