Welcome back

Don't have an account? Sign up

Getting started

An AI Thing lets you chat with 20+ AI models, build reusable agents, and chain them into multi-step workflows — no code required.

[placeholder — short intro video (60–90 s) showing the full loop: open the app, start a chat, create an agent, run a workflow]

Pick a plan on the Pricing page and subscribe.
Start a chat — go to Chats and press New chat to talk to any model immediately.
Create an agent — go to Agents, press New agent, give it a name and a system prompt, then save.
Build a workflow — go to Workflows, press New workflow, add steps that chain agents together.

[placeholder — screenshot of the dashboard / navigation bar highlighting Chats, Agents, Workflows, Compare, Workspaces]

Chats

Chats are multi-turn conversations with any model, agent, or workflow. Each chat keeps the full history so the AI retains context across messages.

[placeholder — screenshot of the chat interface with a conversation open, showing the message input and model selector]

Starting a chat

Go to Chats → New chat. Choose a model (or an agent/workflow) from the selector at the top, then type your message and press Send or hit Enter.

Switching models mid-conversation

Use the model selector dropdown at the top of any open chat. The conversation history is preserved; the next reply will come from the newly selected model.

Attaching files (RAG)

Click the attachment icon to upload a PDF, DOCX, TXT, or image file. The content is indexed and used as context for your messages — useful for summarising documents or asking questions about a file.

[placeholder — screenshot showing the attachment icon and a file already attached to a chat]

Exporting a chat

Open the chat options menu (⋯) and choose Export. You'll receive a ZIP archive with all messages and attachments.

Agents

Agents are reusable AI personas with a fixed model, system prompt, and parameter settings. Create one agent per task — a copywriter, a code reviewer, a customer support bot — and reuse it across chats and workflows.

[placeholder — screenshot of the Agent create/edit form showing the name, system prompt, model selector, and temperature slider]

Creating an agent

Go to Agents → New agent.
Give it a name and an optional cover image.
Write a system prompt — this is the instruction that shapes the agent's behaviour for every message.
Select a model and tune Temperature (creativity) and Max tokens (response length).
Press Save.

Using the Promptify assistant

Not sure how to write a system prompt? Click Promptify next to the prompt field and describe what you want the agent to do in plain English. An AI will draft the system prompt for you.

[placeholder — short video (30–45 s) showing the Promptify button being clicked and the generated system prompt appearing]

Running an agent

From the agent detail page, click Try in playground to test it with a single message. Click Start chat to open a full multi-turn conversation with that agent.

[placeholder — screenshot of the agent playground with a test message and the agent's response]

Workflows

Workflows chain agents into sequential pipelines. Each step receives the output of the previous step, letting you automate multi-stage tasks — e.g. Research → Summarise → Translate.

[placeholder — diagram or screenshot showing a three-step workflow with arrows between agents]

Creating a workflow

Go to Workflows → New workflow and give it a name.
Click Add step, choose an agent, and write its input template.
Repeat for each step. Use {{input}} to reference the workflow's initial input and {{prev_output}} to reference the previous step's result.
Press Save.

[placeholder — screenshot of the workflow step editor showing the template input field with in use]

Running a workflow

From the workflow detail page, type your initial input and click Run. Each step runs in sequence and you can watch the output appear in real time.

[placeholder — short video (30–45 s) of a workflow running step by step with streaming output]

Template variables

{{input}} — the text you typed when starting the run
{{prev_output}} — the full text output from the previous step

Compare

The Compare view lets you run the same prompt against multiple models side by side so you can evaluate quality and speed differences at a glance.

[placeholder — screenshot of the Compare page with two or three models showing different responses to the same prompt]

Running a comparison

Go to Compare.
Select two or more models from the model picker.
Type your prompt and click Run. All responses stream in simultaneously.

Saving comparisons

After a run, click Save to store the comparison. Saved comparisons appear in your Compare history for future reference.

Workspaces

Workspaces (Team plan) let you share chats, agents, and workflows with colleagues under a shared dashboard. Each workspace has role-based access control.

[placeholder — screenshot of a workspace dashboard listing shared chats and agents with member avatars visible]

Creating a workspace

Go to Workspaces → New workspace.
Give it a name, then press Create.
Click Invite members and enter their email addresses.

Roles

Owner — full control: invite/remove members, edit and delete the workspace.
Editor — can create and edit chats, agents, and workflows inside the workspace.
Viewer — read-only access to all shared resources.

[placeholder — screenshot of the workspace members panel showing the three role options in the role dropdown]

Plans & billing

[placeholder — screenshot of the Pricing page comparing the three plans side by side]

Available plans

Personal ($20/mo) — chats, models, and agents for individuals.
Professional ($30/mo) — adds workflows and model comparison.
Team ($40/mo) — adds shared workspaces and multi-user collaboration.

Managing your subscription

Go to Settings → Billing to view your current plan, update payment details, or cancel. All billing is handled securely via Stripe.

Upgrading or downgrading

Visit the Pricing page and select a new plan. Changes take effect immediately; billing is prorated.

FAQ

Which AI models are available?: We support 20+ models from Anthropic (Claude), OpenAI (GPT, o-series), Google (Gemini), DeepSeek, xAI (Grok), Moonshot, and more. Visit the Models page for the full list.
Can I use my own API keys?: No — all model calls go through our platform. Your subscription covers usage; you don't need separate API keys.
Is there a usage limit?: Reasonable use is included with every plan. Very high-volume usage may be subject to rate limits.
What file types can I attach to a chat?: PDF, DOCX, TXT, most plain-text code files, and common image formats (PNG, JPG, WEBP). The content is extracted and used as context for your messages.
Are my chats private?: Yes. Your personal chats are only visible to you. Workspace chats are visible to workspace members based on their role.
How do I cancel my subscription?: Go to Settings → Billing and click Manage subscription. You can cancel at any time; access continues until the end of the billing period.
I found a bug or have a feature request.: Email us directly at support[at]anaithing.com. We read everything.

[placeholder — screenshot or short video showing how to reach the Settings → Billing page and the Manage subscription button]