About TwinGP AI
TwinGP AIIncluded for usersCost saving
TwinGP gives you the AI — you do not need to buy a separate frontier API subscription just to start chatting, talking with the avatar, or running agent and automation AI steps on your plan.
Your workspace uses TwinGP AI with the AI token budget included in Free / Starter / Pro / Enterprise. Open TwinGP and start working — model access is part of the product.
Big model. Low cost. Why?
For every small task, you should not waste money running a full “always-on big model” as if every message needed maximum firepower.
TwinGP AI is still a big model — capable for hard work — but the cost stays low. Why?
Because TwinGP AI is an MoE (Mixture of Experts) model with smart activation:
- It does not burn the whole heavy stack for every tiny ask
- It activates the right expert at the right time
- Simple work stays cheap; complex work wakes stronger experts when needed
- You get big-model quality when it matters, without big-model waste on every click
That is the TwinGP cost story: big model + MoE smart activation = cost saving for real daily use.
Why TwinGP AI is cost-saving
| Idea | What it means for you |
|---|---|
| Don’t waste money on every task | Not every chat, classify, or draft should bill like a maxed-out frontier call |
| Big model, still cost-low | TwinGP AI remains a powerful big model — MoE keeps spend efficient |
| Smart activation (MoE) | Right expert activates at the right time; idle experts stay quiet |
| Plan tokens stretch further | Chat, Talk, Agents, and Automation stay affordable on included tokens |
| One product bill | AI usage rides your TwinGP plan instead of juggling many vendor invoices on day one |
In short: no waste on every task → smart MoE activation → big model, low cost.
Where TwinGP AI is used
| Surface | How TwinGP AI helps |
|---|---|
| Chat and Build | Describe outcomes; TwinGP AI helps create agents, pieces, and automations |
| Talk with AI | Face-to-face live avatar conversation powered by TwinGP AI (or your agent brain) |
| AI Agents | Reasoning and drafting with tools and knowledge |
| Agentic AI Automation | Classify, draft, and decide inside workflow steps |
| Agentic AI | Multi-step plan → act → check → retry intelligence |
What you get on your plan
- Free — try TwinGP with a one-time AI token grant (then upgrade to continue)
- Starter / Pro — monthly AI token budget that resets each paid month
- Enterprise — custom TwinGP AI Agent + flow token capacity
Details: Plans.
Check remaining usage anytime under Settings → Usage (and Plan).
BYOK (optional — advanced)
TwinGP already provides the model for normal users.
BYOK (Bring Your Own Key) is optional when your plan and workspace allow it: connect your own provider key for specific agent or automation steps if your company policy requires it. Most users should stay on TwinGP AI + plan tokens for the simplest cost-saving path.
User tip
- Prefer TwinGP AI for Chat, Talk, Agents, and Automation.
- Watch AI tokens on your plan so usage stays predictable.
- Use BYOK only if you need your own provider — not because TwinGP lacks a model.
TwinGP AI = included big model, MoE smart activation, you don’t waste money on every task, cost stays low.