ThinkLLM
A knowledge graph of AI models
Last verified:
What is ThinkLLM?
ThinkLLM is a living, browsable knowledge graph about AI models that helps users find the right model for their project. The platform covers 1,500+ AI models, connecting them with capabilities, use cases, glossary terms, research papers, and other useful resources. Users can browse by task, compare capabilities, and build model intuition about which models exist, what they're good at, and how they compare.
Key features include browsing by capability or use case, comparing model capabilities across different dimensions, viewing detailed model pages with capability scores, accessing a glossary of AI terms, and exploring connections between models and their applications. The knowledge graph covers both open-weight models that can run locally and commercial API models from major providers. Model descriptions, capability scores, and glossary definitions are AI-generated content produced by language models based on publicly available data.
ThinkLLM is designed for developers, researchers, Enterprise Architects, and anyone curious about AI who needs to make informed decisions about LLM choices. It's particularly useful for those working with AI models in production, selecting models for specific tasks like coding, creative writing, customer support, data analysis, or running models on local hardware for privacy and cost reasons.
The platform is maintained by AI agents that continuously discover new models, gather metadata from publicly available sources, and synthesize original descriptions and capability scores. Every entry is reviewed before publication to ensure quality and accuracy.
ThinkLLM pricing
Pricing model: Freemium
Free - ThinkLLM is completely free to access with no paid tiers or paywalls. The site provides informational content about AI models without any subscription required.
ThinkLLM pros
- Covers 1,500+ AI models in one place
- Browse by task or use case for quick model discovery
- Compare capabilities across different models
- Includes both open-weight and commercial API models
- AI agents continuously update the knowledge graph
- Every entry is reviewed before publication
- Detailed capability scores for each model
- Built-in glossary of AI terms and concepts
- Connects models to research papers and resources
- Helps build practical model intuition
- Free to access with no paywall
- Search by provider to find models from specific companies
- Shows which models work best for specific tasks
- Original AI-generated descriptions (not copied)
- Good for Enterprise Architects working with capabilities and use cases
ThinkLLM cons
- No API access for programmatic queries
- AI-generated content may contain inaccuracies
- Limited to publicly available model information
- No user accounts or saved comparisons
- Capability scores are AI-generated, not benchmarked
- No community reviews or ratings
- Cannot filter by specific hardware requirements
- No direct links to model APIs or pricing
Frequently asked questions about ThinkLLM
What is ThinkLLM?
ThinkLLM is a living, browsable knowledge graph about AI models. It helps developers, researchers, and anyone curious about AI build practical intuition about the model landscape by showing which models exist, what they're good at, and how they compare.
How many models does ThinkLLM cover?
ThinkLLM covers 1,500+ AI models, including both open-weight models that can run locally and commercial API models from major providers.
How is the data kept up to date?
AI agents continuously discover new models, gather metadata from publicly available sources, and synthesize original descriptions and capability scores. Every entry is reviewed before publication.
Is ThinkLLM free to use?
Yes, ThinkLLM is completely free to access. There are no paid tiers, subscriptions, or paywalls on the site.
Can I browse models by use case?
Yes, you can browse by task or use case. ThinkLLM organizes models for specific tasks like multi-step planning, coding, creative writing, customer support, data analysis, local deployment, RAG, and more.
Are the model descriptions written by humans?
Model descriptions, capability scores, and glossary definitions are AI-generated content - original text produced by language models based on publicly available data. They are not copied from any source, but every entry is reviewed before publication.
Can I compare multiple models side by side?
Yes, ThinkLLM allows you to compare capabilities across different models. You can browse by capability, use case, or provider to see how models stack up against each other.
Does ThinkLLM cover open-weight models?
Yes, the knowledge graph covers both open-weight models (that can run on local hardware for privacy, cost, or offline use) and commercial API models from major providers.
How can I report inaccurate information?
If you believe content on the site infringes your rights or contains inaccurate information about your model or organization, you can contact ThinkLLM at [email protected]. They will review and respond to all requests promptly.
What capabilities can I filter by?
ThinkLLM covers capabilities including multi-step reasoning, code generation, creative writing, world knowledge accuracy, instruction following, long context retention, non-English language quality, vision/audio understanding, function calling, structured output, and agent-style tool orchestration.