Selin Docs

Selin is free to use — no credit card required. Create an account at selin-ai.com/register, then open the chat from your dashboard.

Sending a message: Type in the message box and press Enter (or Shift+Enter for a new line). Selin will respond instantly — she detects your domain and emotional tone automatically.

Starting a new chat: Click New Chat in the sidebar at any time. Your previous conversations are saved and searchable.

Selin offers three model tiers. Select one using the buttons in the chat toolbar — your choice is saved across sessions.

Deep Think enables extended, step-by-step reasoning powered by Selin Ultra. Use it for complex problems that benefit from careful multi-step analysis — proofs, code architecture, legal research, or nuanced strategy.

How to enable: Click the Deep Think button in the chat toolbar (it turns violet when active). Selin will show a collapsible thinking trace before her final answer.

Deep Think uses a 10,000-token reasoning budget and supports up to 16,000 tokens in the response. Responses take longer than standard mode — expect 10–30 seconds for complex problems.

Selin supports voice in two forms: inline voice input (tap the mic in the chat bar to dictate a message) and Voice Mode (hands-free back-and-forth conversation).

Entering Voice Mode: Click Voice Mode in the chat toolbar. Selin will listen, transcribe your speech, respond, and speak her reply aloud — automatically, without you pressing anything.

Text-to-speech: Selin reads responses aloud using natural TTS. You can enable or disable this in Settings → Voice.

Selin can search the web in real time when she needs current information. She decides automatically when to search — you don't need to ask.

When a search is running, you'll see a "Searching…" indicator in the chat. After the search, Selin cites her sources inline so you can verify any fact.

When search is used: News, recent events, prices, live data, documentation for specific library versions, and any time Selin determines her training data may be outdated.

Attach files to any message using the paperclip icon in the chat bar. Selin will read and reason over the file contents.

Supported formats:

• Documents: PDF, DOCX, DOC, TXT, MD
• Spreadsheets: XLSX, XLS, CSV
• Code: JS, TS, PY, HTML, CSS, JSON
• Images: JPG, PNG, GIF, WebP (vision analysis)

Maximum file size: 10 MB. For images, Selin will describe, analyse, or extract text depending on your prompt.

Projects let you group related conversations under a shared context. Create a project from the sidebar, then open conversations inside it. You can give each project a system prompt that Selin reads before every message in that project.

Memory is automatic. As you chat, Selin extracts and stores key facts about you — your preferences, goals, role, and ongoing work. These memories are injected into every conversation so Selin always knows who she's talking to.

View, edit, or delete your memories from Settings → Memory in the chat sidebar.

Canvas is a persistent document editor with AI assist built right in. Navigate to /canvas from the dashboard to create and manage your documents.

Writing: Type freely — Canvas autosaves every 1.5 seconds. Documents are stored to your account and accessible from any device.

Ask Selin: Select any text, then open the Ask Selin panel. Choose a quick action (Improve, Shorten, Expand, Fix Grammar, Translate, Summarize…) or write a custom prompt. Selin streams her suggestion — you can apply it or discard it.

Build your own AI persona from /assistants. Each assistant has a name, icon, description, base persona, and custom instructions.

Custom instructions are injected at the top of every conversation with that assistant — use them to define tone, rules, response style, or domain focus.

Base persona layers one of Selin's built-in domain modes (Legal, Medical, Developer, Coach, Analyst, Creative) on top of your instructions.

Assistants can be made public so other Selin users can discover and use them.

Tell Selin to remind you about anything, and she'll schedule it automatically — no form to fill out.

"Remind me to submit the report at 9am tomorrow"
"Set a reminder for Friday afternoon to review the contract"
"Remind me in 2 hours to take a break"

Selin extracts the time, date, and context from natural language and creates the reminder in the background. You'll see a confirmation in the chat when it's set.

View and manage all your reminders from Settings → Reminders in the chat sidebar.

Personas tune Selin's tone, vocabulary, and depth for a specific domain. Select one from Settings → Persona in the chat sidebar.

Generate API keys from /keys. Keys are used to authenticate requests to the Selin API.

Key limits by plan:

• Free — 2 API keys
• Pro — 6 API keys
• Elite — 10 API keys
• Enterprise — Unlimited

Each key can be named and deactivated individually. Treat your keys like passwords — do not commit them to source control.

Authorization: Bearer sk_selin_xxxxxxxxxxxxxxxxxxxx

See the API Reference for full endpoint documentation.

Selin Code is a terminal-native AI assistant with tool use, git integration, and code execution. It runs as an autonomous agent — give it a task in plain English and it reads files, edits code, and runs commands.

Installation:

npm install -g selin-code

Authentication:

selin auth

Start the REPL:

selin

Select a model tier at startup:

selin --tier spark    # fast, everyday tasks
selin --tier core     # balanced (default)
selin --tier ultra    # maximum reasoning

Key commands:

/add src/app.js       # load a file into context
/glob src/**/*.ts     # load files by pattern
/review               # code review with suggestions
/debug                # debug the code in context
/diff                 # analyse git diff
/commit               # generate a commit message
/tier ultra           # switch model tier mid-session
/help                 # show all commands

Selin offers four plans:

Manage your subscription, upgrade, or cancel from /billing. Pro includes a 7-day free trial — no charge if you cancel before the trial ends.