NovAI — AI for Discord

Quick Start

3 steps to your first generation

01 — INSTALL

Add NovAI to your server

Click the invite link and authorize NovAI with the required permissions. Works in servers and DMs.
→ Add to Discord

02 — GENERATE

Try your first command

No sign-up needed. Type any of these right away. /ask what is the speed of light /image a fox in a forest

03 — UNLOCK (OPTIONAL)

Connect your account

To access paid models (Claude, Gemini, Veo…) link your Pollinations account for free. /connect

Commands

● free | ● account required

💬 Chat & Text

/text [prompt] free

Opens a persistent AI chat thread. Keep typing in the thread to continue — no re-typing the command.

/ask [prompt] free

Get a quick reply inline, no thread. Best for one-off questions.

/privtext [prompt] free

Opens a private thread — only you and NovAI can see it. Great for sensitive topics.

/privchat free

Same as privtext, but no opening message needed. Just start chatting.

/translate [text] [lang] free

Translate to any language. Supports 100+ languages including Japanese, Arabic, Swahili.

/summarize [text] [style] free

Summarize in 4 styles: bullet points, paragraph, one sentence, or TL;DR.

🖼️ Image

/image [prompt] free

Generate an image. Supports 3 sizes: square, landscape, portrait. Includes a copy-URL button.

/batch [prompt] [2-4] free

Generate 2–4 image variations in parallel. All generated with different seeds.

/edit [image_url] [prompt] account

Edit an existing image using FLUX.1 Kontext. Paste any image URL and describe the change.

🔊 Audio & 🎬 Video

/audio [text] free

Convert text to speech. 11 OpenAI TTS voices available for free (Nova, Alloy, Echo…).

/video [prompt] account

Generate a short video clip using Veo 3.1 or Seedance. Requires Pollen credits.

⚙️ Settings & Models

/model [provider] [type] [name] account

Switch provider and model for text, image, audio, or video. Full autocomplete in Discord — e.g. /model provider:llm7 type:text elixosearch

/models [provider] free

List all available models for a provider. Shows free models without an account, all models with one.

/info free

View your current settings: active provider, models, connected keys, and saved memory.

/connect [provider] free

Link any provider — e.g. /connect provider:sixfinger. Pollinations uses free device-flow, SixFinger opens a secure key modal.

/disconnect [provider] free

Unlink a provider — e.g. /disconnect provider:gemini. Keys are removed from memory instantly.

/profile account

See your Pollinations tier, GitHub username, and Pollen credit balance.

🧠 Memory & 🛠️ Utilities

/remember [key] [value] free

Tell NovAI something to remember — your name, language, preferred tone. Persists across threads.

/forget free

Clear everything NovAI has stored about you.

/poll [topic] free

AI generates a poll question with 4 choices and auto-adds 🇦🇧🇨🇩 reactions.

/roast [target] free

Get a spicy but harmless comedy roast. Pure wit, no slurs. Good for server banter.

/export free

Download your current thread's chat history as a .txt file.

/reset · /ping · /help free

Close the current thread, check latency with a color bar, or see the full command reference.

Models

🌸 Pollinations models shown · Yellow = Pollen credits

💬 Text

🖼️ Image

🔊 Audio

🎬 Video

GPT-5.4 Nanofree

GPT-5.4 Minifree

GPT-5.5free

DeepSeek V4 Flashfree

DeepSeek V4 Profree

Mistral Small 4free

Mistral Large 3free

Meta Llama 4 Scoutfree

Qwen3 Coder 30Bfree

Grok 4.20free

Grok 4.3free

Perplexity Sonarfree

Moonshot Kimi K2.6free

MiniMax M3free

Nova 2 Litefree

Z.ai GLM-5.2free

Gemini 3.5 Flashpaid

Gemini 3.1 Propaid

Gemini 2.5 Flash Searchpaid

Claude Haiku 4.5paid

Claude Sonnet 4.6paid

Claude Opus 4.7paid

Claude Opus 4.8paid

Mercury 2paid

Qwen3 Coder Nextpaid

Meta Llama 4 Maverickpaid

Default: GPT-5.4 Nano · Use /model to switch · /models for full list

Flux Schnellfree

FLUX.2 Klein 4Bfree

FLUX.1 Kontextfree

GPT Image 1 Minifree

GPT Image 1.5free

Z-Image Turbofree

Nova Canvasfree

Seedream 4.0paid

Seedream 4.5 Propaid

Seedream 5.0 Litepaid

Ideogram 4.0 Turbopaid

Ideogram 4.0 Qualitypaid

Grok Imaginepaid

Grok Imagine Propaid

GPT Image 2paid

Wan 2.7 Imagepaid

Default: Flux Schnell · /edit uses FLUX.1 Kontext

Nova / Alloy / Echofree

Fable / Onyx / Shimmerfree

Ash / Ballad / Coralfree

Sage / Versefree

Whisper Large V3free

AssemblyAI Universal-2free

ACE-Step 1.5 Turbofree

ElevenLabs v3 TTSpaid

ElevenLabs Flash v2.5paid

ElevenLabs Multilingual v2paid

ElevenLabs Musicpaid

ElevenLabs Sound Effectspaid

Qwen3-TTS Flashpaid

Stable Audio 2.5paid

Stable Audio 3paid

Scribe v2paid

Default: Nova · 11 free OpenAI TTS voices · Paid models require /connect

LTX-2.3free

Nova Reelfree

Veo 3.1 Fastpaid

Seedance Pro-Fastpaid

Seedance 2.0paid

Wan 2.7paid

Wan 2.7 1080ppaid

Grok Video Propaid

Default: Veo 3.1 Fast · All video requires /connect and Pollen credits

Features

🧠

Persistent Memory

Tell NovAI your name, preferred language, or writing style with /remember and it'll apply across every conversation.

🔒

Private Threads

Use /privtext to spin up a Discord private thread — only you and NovAI can see it. Moderators are excluded by default.

💬

DM Support

Add NovAI to your user account and chat in DMs — no server required. The conversation continues as you keep typing, no slash commands needed.

🎭

8 Built-in Personas

Switch NovAI's personality mid-conversation: default, sarcastic, formal, pirate, anime, hacker, Shakespeare, or chef. Use /text system: to set any custom persona too.

🖼️

Batch Image Generation

Generate up to 4 image variations in parallel from a single prompt — all with different seeds. Each gets its own URL copy button.

🌐

BYOP — Bring Your Own Pollen

Your Pollen credits pay for your own usage — NovAI never touches your money. Link via /connect using secure device flow.

📋

Chat Export

Download any thread's full conversation history as a clean .txt file with /export.

⚡

Autocomplete Everywhere

Model names, types, and options all autocomplete as you type in Discord. No need to remember exact model strings.

Access

No account needed

Free tier

16+ free text models (GPT-5.4 Nano default)
Flux Schnell image generation
11 OpenAI TTS voices (Nova, Alloy, Echo…)
Batch generation up to 4 images
DM conversations, private threads
Memory, personas, translate, summarize
No sign-up, no API key

With Pollinations account · free to create

Full access

All 50+ models across text, image, audio, video
Claude Opus, Gemini Pro, Grok, and all paid models
Video generation (Veo 3.1, Seedance, Wan)
ElevenLabs TTS, Stable Audio, music generation
Image editing via FLUX.1 Kontext
Pollen credits: $1 ≈ 1 Pollen, credits never expire
Your usage, your Pollen — NovAI earns a small referral fee

FAQ

Is NovAI free to use?↓

Yes — 16+ text models, image generation, and 11 TTS voices are completely free with no sign-up. Paid models (Claude, Gemini, Veo…) require a free Pollinations account and Pollen credits ($1 ≈ 1 Pollen, never expire).

Is my data stored anywhere?↓

NovAI stores conversation history and memory only in-memory while the bot is running — nothing is persisted to a database. Your Pollinations API key is also only stored in memory and is never logged. Restart the bot and everything resets.

How does /connect work? Is it safe?↓

NovAI uses Pollinations' secure OAuth device flow — you get a one-time code, enter it at enter.pollinations.ai/device, and authorize NovAI. Your sk_ key is stored in memory only and masked in all logs. NovAI never sees your password.

Can I use NovAI without a server?↓

Yes. NovAI supports User Install — add it to your Discord account and use it in any DM or group DM. Just type a message after using /text or /privchat and the conversation continues automatically without re-invoking any commands.

How do I self-host NovAI?↓

Clone the GitHub repo, create a .env with your DISCORD_TOKEN, install dependencies with pip install discord.py aiohttp python-dotenv, and run python Bot.py. Deploy on Railway, Fly.io, or any VPS for 24/7 uptime.

What's the difference between /text and /ask?↓

/ask sends a single question and gets a single reply inline — no thread. /text opens a dedicated thread where you keep chatting by just typing, with full conversation history maintained. Use /ask for quick lookups, /text for actual back-and-forth.