Router One subscriptions are live. Alongside our pay-as-you-go wallet, you can now subscribe to a monthly plan that gives you unlimited access to a curated set of models — no per-token tracking for the models you use most.
Six plans launch today across two product lines and three tiers. Pick the one that matches your workflow, or stack subscriptions with your existing wallet balance. Browse them all on the pricing page.
Two product lines
We split plans by how you primarily work with models:
- Coding — for editor agents and CLI tools like Claude Code, Codex, and Cline. Unlimited access to the coding models you actually run through agent loops.
- Chat — for research, writing, document work, and general Q&A. Unlimited access to the chat models you reach for in a browser tab or a desktop app.
The two are fully independent. If you write code during the day and use chat for research at night, subscribe to both — they activate in parallel and don't share quota.
Three tiers
Each product line has three tiers — Pro, Max, and Ultra — that differ in which models are unlimited and how priority is allocated under load.
| Plan | Monthly | Unlimited models | Premium quota |
|---|---|---|---|
| Ai Coding Pro | $20.99 | Standard coding models (deepseek-chat, deepseek-coder, gemini-2.5-flash) | $5 |
| Ai Coding Max | $47.99 | Workhorse coding models (adds claude-sonnet-4-6, gpt-5.4) | — |
| Ai Coding Ultra | $149.99 | Premium coding models (adds claude-opus-4-7, gpt-5.5) | — |
| AI Chat Pro | $9.99 | Standard chat models (deepseek-chat, gemini-2.5-flash) | $3 |
| AI Chat Max | $29.99 | Workhorse chat models (adds claude-sonnet-4-6, gpt-5.4) | — |
| AI Chat Ultra | $99.99 | Premium chat models (adds claude-opus-4-7, gpt-5.5) | — |
All plans run for 30 days from purchase. No auto-renew, no surprise charges.
How plans work
Unlimited within fair use. Each plan lists exactly which models are unlimited. Calls to those models don't draw down your wallet — they're fully covered for the duration of the subscription, within fair-use limits.
Premium quota on Pro tiers. Pro plans include a small monthly USD quota ($5 for Coding, $3 for Chat) that you can spend on any model not in the unlimited list. Useful when you occasionally need a heavier model without upgrading the whole subscription.
Wallet balance stays untouched. Subscriptions don't replace pay-as-you-go. Your existing wallet balance keeps working for any model outside your subscription's unlimited list, with no double-billing.
Stackable and queueable. Buy a second copy of the same plan and it queues to start the moment your current one expires — zero overlap, zero gap. Buy a different plan (e.g., Coding Max + Chat Ultra) and both run in parallel.
Picking a tier
Three quick decision rules:
- Coding or Chat? Pick by where you spend most of your tokens. You can always add the other later.
- Run agents daily? Skip Pro, go straight to Max. The extra coverage of
claude-sonnet-4-6andgpt-5.4pays for itself within a few sessions of heavy use. - Long contexts, complex tasks, or production agents? Ultra. Highest fair-use priority and the broadest unlimited lineup including
claude-opus-4-7.
If you're not sure, start with Pro and upgrade after 30 days. Subscriptions don't lock you in beyond the current cycle.
Fair use and priority
All plans operate under fair-use limits, with three priority levels:
- Standard (Pro) — typical priority queue
- Priority (Max) — preferred routing during peak hours
- Highest priority (Ultra) — front of the queue, lowest queueing latency
This matters during traffic spikes. If you're running production agents and need consistent latency, Ultra's priority is the practical reason to upgrade.
How to subscribe
Open the pricing page, pick a plan, and pay with your wallet balance (instant activation), Alipay, or crypto. Balance payment is the fastest — the subscription is active before the dialog closes.
Manage your active subscriptions and queued renewals on the subscription dashboard.
Questions or feedback? Find us on Telegram.