A URL with a tracking code that tells a vendor which site or person referred a sale. When you click an affiliate link on this directory and buy a tool, we earn a commission. The tool price is the same for you either way. Every listing discloses whether an affiliate relationship exists.
AI agent
A program that uses an AI model to plan and carry out multi-step tasks on its own - browsing the web, calling APIs, writing and running code - rather than just answering a single question. Most tools marketed as 'AI agents' are closer to complex automations; true autonomous agents that reliably handle open-ended tasks are still emerging.
API (Application Programming Interface)
A standardised way for two pieces of software to communicate. When Zapier connects your Etsy shop to your email platform, it is using each platform's API. You rarely interact with APIs directly as a solo seller, but they power every integration and automation tool in this directory.
C
Context window
The maximum amount of text an AI model can read and respond to in a single conversation. Measured in tokens (roughly 3/4 of a word each). A 200k-token context window can hold roughly 150,000 words - about two full novels. Short context windows mean the model forgets early parts of a long conversation.
D
Diffusion model
The type of AI model that powers most image generators including Midjourney, Stable Diffusion, and DALL-E. It works by learning to remove noise from images iteratively, effectively 'dreaming up' images that match a text description. Most AI product mockup tools are built on diffusion models.
E
Embedding
A way of representing text (or images) as a list of numbers that captures meaning. Words or phrases with similar meanings end up close together in this numerical space. Embeddings power search, recommendations, and tools like Chatbase that find the most relevant part of your policy document to answer a buyer question.
F
Fine-tuning
Training a pre-existing AI model further on your own specific data so it behaves differently - writing in your brand voice, answering only about your product range, or following your shop policies. More powerful than prompting alone but requires technical setup and enough examples. Most small sellers use prompting rather than fine-tuning.
Freemium
A pricing model where the core product is free but advanced features, higher usage limits, or commercial rights require a paid subscription. Most tools in this directory are freemium. The important question is not whether there is a free tier, but whether the free tier is useful enough to evaluate the tool properly before paying.
G
Generative AI
AI that creates new content - text, images, audio, or video - rather than only classifying or analysing existing content. ChatGPT, Midjourney, and Canva's AI features are all generative AI. The term is often used interchangeably with 'AI' in current conversation but is technically more specific.
H
Hallucination
When an AI model confidently states something that is false or made up. A common failure mode: asking ChatGPT for competitor pricing may produce plausible-sounding numbers that are entirely wrong. Always verify factual claims - especially prices, statistics, or product specifications - from a primary source.
I
Inference
The process of running an AI model to generate an output - a response, an image, a prediction. You are paying for inference whenever you use a per-message or per-image pricing model. The cost of inference has dropped dramatically since 2022, which is why AI tools have become affordable for solo sellers.
L
Large Language Model (LLM)
An AI model trained on a large body of text that can generate, summarise, translate, and reason about language. GPT-4o (ChatGPT), Claude, and Gemini are all LLMs. They are the core technology behind AI writing assistants, chatbots, and most AI tools that work with text.
M
Multimodal
An AI model or tool that can work with more than one type of input - text, images, audio, or video. GPT-4o and Gemini are multimodal: you can paste a photo of your product and ask for listing copy describing it. Multimodal capability makes AI assistants considerably more useful for product sellers.
N
n8n
An open-source workflow automation platform (listed in this directory) that you can self-host for near-zero cost. Unlike Zapier or Make, which charge per operation, self-hosted n8n has no per-run fees. Requires basic technical confidence to set up and maintain.
P
Prompt
The instruction or question you give to an AI model. A well-written prompt includes context (who you are, what you sell), a clear task (write a title for this product), constraints (under 140 characters, for Etsy search), and optionally an example of the style you want. Better prompts produce dramatically better output.
Prompt engineering
The practice of writing, testing, and refining prompts to get consistently better outputs from an AI model. Not as technical as it sounds - it mostly means being specific, providing context, and iterating. Experienced digital sellers often maintain a library of tested prompts for their most common tasks.
R
RAG (Retrieval-Augmented Generation)
A technique where an AI model is given relevant documents to read before answering, rather than relying only on its training data. This is how tools like Chatbase work: your shop FAQ is stored as embeddings, and when a buyer asks a question, the relevant FAQ section is retrieved and fed to the model to generate an accurate answer.
S
Stable Diffusion
An open-source image generation model that powers many AI image tools, including some commercial ones. Because the model weights are freely available, developers have built thousands of fine-tuned variants for specific styles. Leonardo AI and some features of other tools use Stable Diffusion variants under the hood.
System prompt
Instructions given to an AI model before the conversation starts that shape how it responds throughout. Custom GPTs in ChatGPT, and many AI tools built for specific niches, use system prompts to pre-load your context - your brand voice, your product catalogue, your customer service policies - so you do not have to repeat it every session.
T
Token
The unit AI models use to measure text. One token is roughly 3/4 of a word in English. 'Etsy seller' is about 3 tokens. Models have a maximum token limit per conversation (the context window), and API pricing is typically quoted per 1,000 or 1 million tokens. For consumer tools like ChatGPT, you do not need to count tokens directly.
V
Vector database
A database designed to store and search embeddings (numerical representations of text or images). Used in tools like Chatbase to find the document sections most relevant to a buyer's question. The technology behind semantic search - finding results that match the meaning of a query, not just the exact words.
W
Webhook
A way for one web service to instantly notify another when something happens. When an Etsy order is placed, Etsy can send a webhook to Zapier, which then triggers your automation. Webhooks are faster and more reliable than polling (repeatedly asking 'has anything changed yet?'). Most automation tools support webhooks.
Workflow automation
Connecting apps and services so they perform repetitive tasks automatically when triggered by an event. Sending a thank-you email after every sale, logging orders to a spreadsheet, or posting to Pinterest when you add a listing are all workflow automations. Tools like Zapier and Make let you build these without writing code.
Z
Zero-shot
When an AI model completes a task from a description alone, with no examples provided. Contrasted with few-shot, where you include 1-5 examples of the output you want. For listing copy, giving the model 2-3 examples of your best existing listings (few-shot) consistently outperforms describing what you want without examples (zero-shot).