Chat, voice, and agent tools tuned for builders — ask in Kurdish or English and get a direct answer, not a lecture.
Not a demo. A working chat, voice, and agent stack tuned for how you actually build.
Debug endpoints, review architecture, ship features — without the model second-guessing your stack.
Auto-detects Sorani or English per message and lays the UI out RTL or LTR to match.
Call in and talk through a problem live — same context carries over into the chat thread.
Token-by-token output in a composer that keeps model picker and history one tap away.
Threads persist to your account — pick a build back up without re-explaining the context.
Start on VX Mini or VX Normal. VX Pro and VX Pro Max add the agent, VX-843B, thinking, and every language including Kurdish Sorani.
Tuned for code, Kurdish translation, and long technical threads. Keeps context across sessions and powers chat, voice, and the agent workspace.
No invite list. Confirm your account once and you're in.
Pick a model, type in the composer, and stream your first answer.
VX Pro and VX Pro Max unlock the agent, thinking style on VX-843B, and a bigger monthly token cap.
Billed once a month. Token limits reset every 30 days. Geo /location (photo → address) is separate: Mini 0 · Normal 1/month · Pro 5/week · Pro Max unlimited.
Checkout total 8,288 IQD (incl. Wayl fee 788)
Checkout total 14,438 IQD (incl. Wayl fee 938)
Checkout total 29,813 IQD (incl. Wayl fee 1,313)
Checkout total 60,563 IQD (incl. Wayl fee 2,063)
You must sign in with Google before you can pay — checkout is tied to your account. Charged in IQD at 1,500 IQD per $1. Checkout total = package + Wayl fee (2.5% + 600 IQD local). That fee is added on our payment link so you pay it with the plan.
Sign in with Google to buy VX. After login we open checkout for your account.
Sign in to payTap Get started, sign in with Google, and complete the one-time onboarding. Paid plans unlock higher monthly usage.
Yes. Every plan speaks Kurdish Sorani. VX Pro and VX Pro Max also handle any other language you type, with the UI switching RTL or LTR to match.
Every plan has a token cap (what you type plus what VX replies, including thinking tokens on VX-843B). VX Mini is 583K tokens on VX-320B. VX Normal is 1.5M on VX-320B. VX Pro is 3.4M on VX-320B or 2.1M on VX-843B. VX Pro Max is 7.2M on VX-320B or 4.3M on VX-843B. Pro and Pro Max include a small extra above the base rate. They share one pool — mix models, but VX-843B uses the budget faster. Limits reset each month. Reply counts on the cards assume ~2,000 tokens per turn.
Attach a photo and send /location. Mini has 0 tries. Normal has 1 try per month. Pro has 5 tries per week. Pro Max is unlimited. Tries only count when Geo succeeds.
VX Mini and VX Normal are chat on VX-320B (256K context), with no thinking trace and no VX-843B. VX Pro and VX Pro Max add the agent workspace, VX-843B Super Intelligent (1M context) with thinking style, and replies in any language including Kurdish Sorani.
Every plan is uncensored — 0 content filters. VX-320B (all plans) is 320B with a 256K context window. VX-843B Super Intelligent on Pro and Pro Max is 843B with a 1M context window. Monthly token caps are separate from how much the model can hold in one thread.
Sign in with Google first, pick a plan, then pay in IQD on the checkout page (FIB, FastPay, ZainCash, QiCard, or card). Payment is always tied to your signed-in account. You pay package price + Wayl fee (2.5% + 600 IQD local) in one checkout total — for example VX Normal is 13,500 + 938 = 14,438 IQD.