Background

Large language models (LLMs) are AI systems that process and generate human-like language by training on massive datasets. Models like OpenAI's GPT-4, Anthropic's Claude Sonnet, and Google's Gemini each bring unique strengths to different tasks. Some excel at creative writing, while others shine in technical precision or handling large amounts of context. These variations stem from differences in model size, training data, and architecture.

Co stands out by offering access to a diverse range of models. Unlike platforms that limit users to a single provider's models (like ChatGPT's OpenAI-only selection), Co gives you the flexibility to choose the best model for your needs. While accessing every state-of-the-art model through individual providers could cost $60+ monthly, Co makes multiple models available to you for free.

Selecting a Model

Choosing the right model depends on your specific needs and preferences. The AI landscape evolves rapidly, making today's top performer tomorrow's runner-up. Here's how to make an informed choice:

1. If you're unsure, start with our recommendations

• We pre-populate your first space with great selected general-purpose models. If you're new to AI or unsure where to start, try these options first. You can always explore more models on the Discover page.

• When creating a bot, chose automatic and let us match you with the optimal model based on your use case. You can easily switch models later if needed.

2. Check current benchmarks

• Data-driven decisions start with reliable benchmarks. We recommend:
• LM Arena for interactive comparisons
• HuggingFace's comprehensive benchmark collection for more further research

3. Consider standard models

• Co categorizes models as either standard or premium based on operating costs, not quality. Many standard models like Gemini Flash 2 and Claude 3.7 Haiku deliver excellent performance for everyday tasks. Try them out for everyday tasks!

4. Choose Gemini for long context

• Google's Gemini models lead the industry in handling extensive context. When working with large documents (like 100-page PDFs), Gemini currently offers unmatched performance by a long shot. If you need to handle large amounts of context, Gemini models are the way to go.

5. Explore model details

• Hover over the info icon when selecting a model to learn about its strengths and context window capabilities.

6. Experiment freely

• The best model for you depends on your specific needs. Don't hesitate to try different options to find your perfect match.

Ready to explore? Start creating your first bot today!