AI models

ASKtoAI offers a variety of AI Models that users can leverage.

This approach provides several advantages and allows access to cutting-edge AI models:

Model Diversity:

ASKtoAI Auto: Uses the best AI adapting to the request of the user , available to all users.
Anthropic Claude 3.5 Sonnet, Claude 3 Sonnet and Claude 3 Opus: Advanced models from Anthropic, known for their high-quality text understanding and generation capabilities.
GPT-3.5 Turbo , GPT-4o and GPT-4o mini: OpenAI's renowned models, with GPT-4 representing the state-of-the-art in many AI applications.
Gemini 1.5 Flash and Pro: Google's latest models, offering high performance and versatility.
Meta Llama 2 (7B and 405B): Meta's open-source models, providing powerful and customizable options.

It's important to note that model availability is tied to the user's subscription plan:

Users on the Base or Professional plans have access to ASKtoAI Auto, the default model.
Users on the Growth plan or Enterprise can select from the full range of AI models listed above.

This tiered approach ensures that all users have access to powerful AI capabilities, while providing additional options and flexibility for users with more advanced needs.

This strategy positions ASKtoAI as a comprehensive AI solutions hub, capable of meeting a wide range of user needs and preferences. By maintaining an easy-to-use interface and offering access to the most advanced AI models available in the market (based on the user's plan), ASKtoAI caters to both casual users and power users lenghtalike.

To provide our users with a clear understanding of the performance and capabilities of the various AI models available on ASKtoAI, we've conducted targeted benchmarks focusing on two critical aspects: Output Speed and Reasoning and Knowledge.

Output Speed: This benchmark measures how quickly each model can generate responses, which is crucial for applications requiring real-time or near-real-time interactions.
Reasoning and Knowledge (MMLU): The Massive Multitask Language Understanding (MMLU) benchmark assesses each model's ability to reason and apply knowledge across a wide range of subjects, including science, mathematics, humanities, and more.

The following graphs illustrate the comparative performance of our available models in these two key areas:

These benchmarks offer a visual representation of how each model performs in terms of speed and comprehensive knowledge application. They can help users make informed decisions when selecting the most suitable model for their specific needs, whether it's for rapid response times or deep, multifaceted reasoning capabilities.

It's important to note that while these benchmarks provide valuable insights, the best model for a particular task may vary depending on the specific requirements of each user's project.

Memory and Memory Plus Model Interactions:

ASKtoAI offers two memory features that significantly enhance the contextual understanding and long-term recall of our AI models:

Memory: Our standard memory feature allows models to retain context within a conversation, improving coherence and relevance in responses.
Memory Plus: This advanced feature dramatically expands the context window for all models, enabling them to process and recall much larger amounts of information.

Model-specific Memory Capabilities:

All models on ASKtoAI are compatible with both Memory and Memory Plus features. However, their input capacities may vary, this means that some models can't be used if the length content is higher then the limit token:

ASKtoAI Auto: Utilizes both Memory and Memory Plus for improved contextual understanding.
Claude models: Fully compatible with Memory and Memory Plus, offering enhanced long-term recall.
GPT models: Work well with both Memory and Memory Plus, maintaining context effectively up to 1M tokens.
Meta Llama 2 models: Compatible with Memory and Memory Plus, providing consistent context retention.
Gemini 1.5 Pro: While all models can use Memory Plus, Gemini 1.5 Pro stands out with its exceptional input capacity. It can process up to 5 million tokens of input, making it particularly well-suited for tasks requiring extensive context or large knowledge bases.

The Memory Plus feature enhances the capabilities of all models, allowing for improved handling of tasks that require processing vast amounts of information. This includes analyzing lengthy documents, maintaining context across multiple conversations, or integrating large knowledge bases into responses.

Gemini 1.5 Pro's ability to handle up to 5 million tokens of input makes it especially adept at tasks involving extremely large datasets or contexts. However, all models benefit from Memory Plus, each offering improved performance within their respective token limits.

We encourage users to consider these metrics in the context of their unique use cases and to experiment with different models (as available in their subscription plan) to find the optimal solution for their needs.