Choosing Your AI Model
Ozmoz provides access to a wide range of AI models, each with unique strengths. The model you choose will directly impact the speed, accuracy, and "personality" of your AI-powered interactions. This guide helps you select the best model for your needs, whether for general dictation, complex problem-solving, or creative writing.
Text Generation Models
These models are used for all AI Agent tasks and general AI generation. The choice involves a trade-off between speed, reasoning capability, and context window size (the amount of text the model can "remember").
| Model | Key Strengths | Best For... | Plan |
|---|---|---|---|
llama-3.1-8b-instant |
Fast, Versatile | Quick questions, summarizing short texts, and general assistant tasks where speed is critical. | Free / Pro |
llama3-70b-8192 |
Versatile, Powerful | High-quality content creation, complex instructions, and tasks requiring deep understanding. A great all-rounder. | Free / Pro |
llama-3.3-70b-versatile |
Ingenious, Large Context | Creative writing, brainstorming complex ideas, and tackling novel problems that require out-of-the-box thinking. | Free / Pro |
qwen/qwen3-32b |
Thinking, Reasoning | Logical problem-solving, code generation, and tasks that require step-by-step reasoning. | Free / Pro |
meta-llama/llama-4-scout-17b-16e-instruct |
Multimodal, Efficiency | Analyzing images or screenshots with maximum efficiency and a very high rate limit. Ideal for frequent visual analysis tasks. | Pro |
meta-llama/llama-4-maverick-17b-128e-instruct |
Multimodal, Fast | Quickly understanding and describing the content of images and screenshots. Excellent for rapid visual Q&A. | Pro |
compound-beta |
Web & Code Synthesis | Complex tasks requiring information synthesis from web sources and advanced code generation. | Pro |
Audio Transcription Models
These models are specifically designed to convert your voice into text. The main difference lies in the balance between transcription speed and pinpoint accuracy.
| Model | Key Strength | Best For... | Plan |
|---|---|---|---|
nova-2 |
Versatile | Everyday use. It offers an excellent balance of high speed and great accuracy for general dictation. | Free / Pro |
whisper-large-v3 |
Precise | Maximum accuracy. Best for transcribing technical jargon, difficult accents, or in noisy environments where every word counts. | Pro |
whisper-large-v3-turbo |
Fast | Near-instant transcription with very high accuracy. It's a faster version of Whisper v3, ideal for conversational speed. | Pro |
Tip: Don't hesitate to experiment! The best model often depends on your specific use case. Try a faster model for quick notes and a more powerful one for detailed work to see what fits your workflow best.