Skip to main content

Documentation Index

Fetch the complete documentation index at: https://info.invictai.io/llms.txt

Use this file to discover all available pages before exploring further.

Our platform provides access to cutting-edge Language Learning Models (LLMs) from leading AI providers. Each model is optimized for diverse tasks with unique capabilities—including high context windows, multimodal (vision) input, and integrated tool use. Note that credits represent resource consumption per 1K tokens.

​
Available Models

Below are the enabled models currently available on our platform – sorted by Elo rating:
Model NameProviderElo RatingContext WindowCredits/1K TokensVisionTool Use
Claude 3.7 SonnetAnthropic1394200k2.610❌✅
ChatGPT-4o LatestOpenAI1365128k4.126✅❌
DeepSeek R1DeepSeek1358128k5.600❌❌
o3-miniOpenAI1325200k0.908❌✅
DeepSeek V3DeepSeek1317128k0.630❌❌
Claude 3.5 SonnetAnthropic1283200k2.615✅✅
GPT-4o miniOpenAI1272128k0.112❌✅
GPT-4oOpenAI1265128k2.137✅✅
Mistral LargeMistral AI125132k1.650❌✅
Llama 3.1 70BMeta1247131k0.487❌❌
Mistral SmallMistral AI122232k0.165❌✅

​
Dynamic Model Categories

These dynamic categories are continually updated by our team to deliver optimal performance and efficiency:
CategoryDescription
AutoBalanced performance model optimized for general use and stability. LLM must have tool use enabled to be in this category
SmartestAlways uses the highest-performing available LLM, automatically updating when superior models are released
EfficientCost-optimized model that maintains smart performance while minimizing resource usage
Use these settings for agents that should always run with the latest optimized models.

​
Intelligence Elo Rating

Our Elo rating — based on the LMSYS Chatbot Arena Leaderboard — assesses model performance in natural language tasks. Higher ratings signify greater proficiency in generating precise and context-aware responses. For more details, please visit: LMSYS Chatbot Arena Leaderboard.

​
Credit consumption

Credits quantify resource consumption. Typically, more advanced models consume higher credits per 1K tokens.

​
Vision Capability

Selected models support multimodal input (e.g., images). Models with vision capability are marked with a ✅ in the Vision column.

​
Tool Use Integration

Many models support external API integrations and advanced data operations through built-in tool use features, as indicated by a ✅ in the Tool Use column.