Glossary

​

A

AI (Artificial Intelligence): Technology that enables machines to perform tasks that typically require human intelligence, like understanding language and generating content. AI Model: A specific AI system designed for particular tasks. Different models have different strengths (e.g., GPT-5 Mini for general tasks, Grok 4 Fast for cost-efficient multimodal tasks). Aspect Ratio: The proportional relationship between width and height of an image. Common ratios include 1:1 (square), 16:9 (wide), and 9:16 (tall).

​

B

Batch Generation: Creating multiple variations of an image at once.

​

C

CFG Scale: Classifier Free Guidance scale - a parameter in image generation that controls how closely the AI follows your prompt. Higher values mean stricter adherence to the prompt. Context Window: The maximum number of tokens (words or word pieces) a model can process in a single conversation, including both input and output. Conversation: A chat session with an AI assistant. Can contain multiple messages back and forth.

​

D

Daily Limit: The maximum number of times you can use a feature per day. Limits reset at midnight in your local timezone. Document Processing: The ability to upload and analyze documents (PDFs, images, etc.) to extract information or answer questions.

​

E

Export: Downloading your data from Nexal AI, including conversations, images, and other content.

​

G

Generation: Creating new content (images, text) using AI. Guidance: A parameter that controls how closely the AI follows your prompt when generating images.

​

M

Memory: A tool that allows the AI to save and recall important information about you across conversations. Memories persist across all chats and help personalize your experience. Model: See AI Model.

​

N

Negative Prompt: In image generation, a description of what you don’t want to appear in the output.

​

P

Persona: A pre-configured AI assistant with specific instructions and personality traits for particular tasks. Premium Model: Advanced AI models that provide better quality and accuracy but may have usage limits. Examples include GPT-5, Grok 4, and Gemini Pro. Prompt: The text you write to tell the AI what you want. Better prompts = better results! Prompt Engineering: The skill of writing effective prompts to get the best AI responses.

​

R

Rapid Mode: Speed over depth. Makes the AI respond faster by skipping the deep thinking. Type !rapid before your message to use it. Reasoning: How much the AI thinks before answering. Low = fast and simple. High = slow and thorough. Most questions work fine with Medium. Reminder: A mobile tool that schedules notifications for future tasks or events.

​

S

Streaming: When AI responses appear word-by-word in real-time, rather than all at once. Style Preset: Pre-configured settings that apply a specific visual style to generated images (e.g., realistic, anime, sketch). Subscription Plan: Your chosen pricing tier (Free, Lite, Plus, Pro, or Max) that determines your features and limits.

​

T

Text-to-Image: Generating images from text descriptions using AI. Text-to-Speech (TTS): Converting text to spoken audio. Used in voice features to have the AI speak responses. Thinking Mode: An advanced chat mode that shows the AI’s reasoning process for complex problems. Tool: A capability that extends the AI beyond text generation, such as web search, image generation, or document creation. Token: A unit of text that AI models process. Roughly equivalent to a word or part of a word.

​

V

Vision: The ability of an AI model to understand and analyze images. Models with vision can describe pictures, read text in images, and answer questions about visual content. Voice Mode: A mobile feature that allows hands-free conversation with the AI using speech-to-text and text-to-speech.

​

W

Web Search: A tool that allows the AI to search the internet for current information when answering questions.

​

AI Provider Glossary

Claude: A family of AI models developed by Anthropic. Includes Claude 3.5 Haiku and other variants. DeepSeek: AI models from DeepSeek, known for strong reasoning capabilities. Includes DeepSeek R1. Gemini: A family of AI models developed by Google. Includes Gemini 2.0 Flash Lite, Gemini 2.5 Flash Preview, and Gemma models. GPT: AI models developed by OpenAI. Includes GPT-5 Mini, GPT-5 Nano, and GPT-4 variants. Grok: AI models developed by xAI. Includes Grok 4 Fast (multimodal with 2M token context) and Grok 3 Mini. Llama: Open-source AI models from Meta. Includes Llama 3.3 70B and 8B variants. Mistral: AI models from Mistral AI. Includes Mistral Small and other variants. Qwen: AI models from Alibaba. Includes Qwen3 variants.