Two ways to pay, your choice
wylon Token Factory will offer both subscription plans and usage-based pricing.
Pick the plan that fits your scale
Pro
For prototyping and personal projects.
- Basic shared inference quota.
- Standard rate limits.
- Try out the full range of models.
- Priority technical support.
Max
For application builders and serious developers.
- Everything in Pro.
- 6× the 5-hour usage of Pro.
- Higher rate limits.
- Early access to new models and features.
Team / Enterprise
For enterprises and teams.
- Shared token quota across the team.
- Higher rate limits.
- Priority onboarding and flexible scaling.
- Commercial support and SLA guarantees.
Chat, image, and video usage pricing
| Capability | Model | Billing unit | Price | Notes |
|---|---|---|---|---|
| Chat | GLM-5.3 | Per 1M tokens | Cache ¥2 / input ¥8 / output ¥28 | Complex engineering and long-horizon agents |
| DeepSeek-V4.1-Flash | Per 1M tokens | Cache hit ¥0.04 / input ¥2 / output ¥8 | Native multimodal and high-throughput inference | |
| Image | Qwen-Image-2512 | Per image | ¥0.30 | Image generation and text rendering |
| SenseNova-U1.5-8B-MoT | Per image | Limited free access | Image generation, editing, and 2K output | |
| Video | Wan2.2-T2V-A14B | Per second | 480p ¥0.20 / 720p ¥0.40 | Fixed five-second video |
| MiniMax H3 · Image input | Per image | ¥0.20 | Usable as a first frame, last frame, first-and-last-frame pair, or reference image; billed by the number of valid input images | |
| MiniMax H3 · Video input | Per second | ¥0.50 | Billed by the number of valid input seconds | |
| MiniMax H3 · Audio input | Per second | ¥0.00 | Billed by the number of valid input seconds; currently free | |
| MiniMax H3 · Video output | Per second | ¥0.50 | Billed by the number of successfully generated seconds |
Your dashboard or sales quote is authoritative for current availability and pricing.
FAQ
When does the subscription quota reset?
Subscription plans are not live yet; the details at launch will be authoritative. The expected quota reset model is as follows:
Rolling-window mechanism:
- 5-hour window: starts at your first request and resets every 5 hours.
- Weekly limit: starts at your first request and resets every 7 days.
You can see live remaining quota and reset times in the dashboard.
How is token usage billed?
Usage is billed separately by the number of input Tokens, cache-hit Tokens, and output Tokens. Input Tokens served from the system-level cache are itemized separately in your usage breakdown. Billing rates are subject to the pricing page.
Subscription or usage-based - which should I pick?
Subscriptions are a good fit for prototyping, fixed monthly budgets, and small teams. Usage-based pricing suits workloads with volatile traffic, per-project billing, or fine-grained cost attribution.
How do I upgrade my subscription?
In your dashboard, go to Benefits → Subscription, pick the plan you want, and complete checkout. Upgrades are prorated - the unused portion of your current plan is refunded, and the new plan's benefits take effect immediately.
