One plan. Every model. $25/month.
Every seat gets full access to every model and every feature. Need more credits? Buy additional packs that carry over month to month.
Pro Seat
For individuals and teams
20M credits per seat · Credits pool across your team
- 50+ AI models — Claude, GPT, Gemini, DeepSeek, Mistral, and more
- Switch models mid-conversation
- Side-by-side model comparison
- Conversation branching with visual graph
- System prompt library
- Goldfish mode for context-free queries
- Fine-tune model behavior (temperature, sampling, and more)
- Conversations stored locally, never on our servers
- Team admin dashboard and usage tracking
Cancel anytime · No long-term contracts
Need custom terms or volume pricing? Contact us
What's in the box
Every feature, every seat. No upgrades, no add-ons.
Seamless model switching
Plan in Opus, implement in Mistral. Switch models mid-conversation without losing context.
Side-by-side comparison
Send the same prompt to multiple models and compare outputs visually.
Conversation branching
Fork conversations at any point. Navigate branches via visual graph and minimap.
System prompt library
Reusable system prompts that work across every model. Build once, use everywhere.
Goldfish mode
One-off questions with zero context. No history pollution, just a clean answer.
Instant local search
Conversations live on your machine, so search is instant across all chats.
Advanced parameters
Temperature, Top P, and more. Fine-tune model behavior for your exact needs.
Complete privacy
Your data never touches our servers. Per-browser storage you fully control.
Rich chat experience
Markdown, syntax highlighting, code blocks — built for long working sessions.
Team management
Invite your team, share a credit pool, and manage access from one dashboard.
Usage dashboard
See who's using what, track credit consumption by model, stay on budget.
50+ models included
Claude, GPT, Gemini, Llama, Mistral, DeepSeek, Grok, Qwen, and more.
Common questions
What are credits and how do they work?
Credits are a universal currency for AI usage. Every seat gets 20 million credits per month, pooled across your team. Different models consume credits at different rates based on their capabilities — a lightweight model uses very few credits per message, while a premium reasoning model uses significantly more. This means your team can send hundreds of messages a day on everyday models, or be more selective when reaching for the most powerful ones.
How far do 20M credits go?
It depends on which models your team uses. On lightweight models (Gemini Flash, DeepSeek, GPT-4o Mini), 20M credits can cover thousands of messages per month — easily hundreds per day. On standard models (Claude Sonnet, Gemini Pro), you're looking at dozens of solid working sessions per day. Premium reasoning models (Claude Opus, o1) use more credits per message, so those are best saved for when you need serious horsepower. Most teams naturally blend — quick questions go to fast models, complex work goes to premium ones — and 20M credits covers that comfortably.
Can admins control which models the team uses?
Yes. Admins can restrict which models are available to the team from the dashboard. If you want to keep costs predictable, you can limit your team to a set of approved models — or leave everything open and let people choose the right tool for the job. Usage is tracked per person and per model, so you always have visibility.
What happens if we run out of credits?
Each seat gets 20M monthly credits that reset every billing cycle. If your team needs more, you can purchase additional credit packs from inside the app. Purchased credits carry over month to month, so nothing goes to waste — they act as a buffer for heavier months without forcing you to upgrade your plan.
Can I use this as an individual, not a team?
Absolutely. Individual accounts get the exact same features and models. You just won't need the team management tools. Same price, same everything.
Is my data really private?
Yes. Think of Cool Ai as a VPN for your AI interactions — your prompts and responses are never stored on our servers. Conversations live entirely in your browser's local storage, and we only track aggregate credit usage for billing. That said, we recommend keeping personally identifying information out of your prompts as a general best practice when using any AI service.
Can I cancel anytime?
Yes. Cancel your subscription at any time from your account settings. You keep access through the end of your current billing period.
How does model switching work?
You can switch to any model at any point in a conversation with one click. The new model picks up the full conversation context. This lets you plan with a powerful model and implement with a fast, cheap one — and your credits stretch further when you use lighter models for the straightforward parts.