Inspiration, Everywhere.
From web pages to documents and every spark of genius in between—there’s a plan designed to keep up with you.
Pick Your Perfect Fit
Each estimate uses the full monthly allowance on one model, with 2,000 uncached input tokens, 12,000 cached input tokens and 240 output tokens per request.
Extra web searches beyond the included limit cost 8,000 credits each.
Final price may vary depending on your location and will be calculated at checkout.
Models & Credits
| Model / input types | Starter | Plus | Pro | Input credits(1M tokens) | Output credits(1M tokens) |
|---|---|---|---|---|---|
Claude Sonnet 5.5 AnthropicTextImagePDF | StarterNot included | PlusIncluded | ProIncluded | Input credits(1M tokens)2M | Output credits(1M tokens)10M |
Kimi K3.1 Pro MoonshotAITextImageVideo | StarterNot included | PlusIncluded | ProIncluded | Input credits(1M tokens)2M | Output credits(1M tokens)10M |
GPT-6.1 Sol OpenAITextImagePDF | StarterNot included | PlusIncluded | ProIncluded | Input credits(1M tokens)2M+ | Output credits(1M tokens)10M+ |
When input length is ≤ 272,000 tokens, the input rate is 2M credits and the output rate is 10M credits. When input length is > 272,000 tokens, the input rate is 4M credits and the output rate is 15M credits. | |||||
Claude Haiku 5.5 AnthropicTextImagePDF | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.1M+ | Output credits(1M tokens)0.5M+ |
When input length is ≤ 100,000 tokens, the input rate is 0.1M credits and the output rate is 0.5M credits. When input length is > 100,000 tokens, the input rate is 0.5M credits and the output rate is 2.5M credits. | |||||
DeepSeek V4 Pro DeepSeekText | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.66M+ | Output credits(1M tokens)1.98M+ |
Pricing varies by Beijing time (UTC+8): peak rates (2× off-peak) apply on weekdays from 09:00 to 12:00 and 14:00 to 18:00; all other times, including weekends, use off-peak rates. | |||||
DeepSeek V4.1 FlashNo window limit DeepSeekTextImage | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.15M+ | Output credits(1M tokens)0.6M+ |
Pricing varies by Beijing time (UTC+8): peak rates (2× off-peak) apply on weekdays from 09:00 to 12:00 and 14:00 to 18:00; all other times, including weekends, use off-peak rates. | |||||
MiniMax-M3 MiniMaxTextImageVideo | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.3M+ | Output credits(1M tokens)1.2M+ |
When input length is ≤ 512,000 tokens, the input rate is 0.3M credits and the output rate is 1.2M credits. When input length is > 512,000 tokens, the input rate is 0.6M credits and the output rate is 2.4M credits. | |||||
MiniMax-M3.1-FlashNo window limit MiniMaxTextImageVideo | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.15M | Output credits(1M tokens)0.6M |
Kimi K2.6 MoonshotAITextImageVideo | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.95M | Output credits(1M tokens)4M |
GPT-5.6 LunaNo window limit OpenAITextImagePDF | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.2M+ | Output credits(1M tokens)1.2M+ |
When input length is ≤ 272,000 tokens, the input rate is 0.2M credits and the output rate is 1.2M credits. When input length is > 272,000 tokens, the input rate is 0.4M credits and the output rate is 1.8M credits. | |||||
GPT-6 LunaNo window limit OpenAITextImagePDF | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)0.1M+ | Output credits(1M tokens)0.5M+ |
When input length is ≤ 272,000 tokens, the input rate is 0.1M credits and the output rate is 0.5M credits. When input length is > 272,000 tokens, the input rate is 0.2M credits and the output rate is 0.75M credits. | |||||
GPT-6.1 Terra OpenAITextImagePDF | StarterIncluded | PlusIncluded | ProIncluded | Input credits(1M tokens)1M+ | Output credits(1M tokens)5M+ |
When input length is ≤ 272,000 tokens, the input rate is 1M credits and the output rate is 5M credits. When input length is > 272,000 tokens, the input rate is 2M credits and the output rate is 7.5M credits. | |||||
Plus and Pro include the same models. Pro has 3× the monthly credits and short-term quotas of Plus.
“No window limit” models draw only from your credit balance, without 5-hour or 7-day limits.
Prices marked “+” are base rates. Expand for long-context rates or variable-pricing details. Models, prices and applicable limits may change.
Usage Limits
On top of the monthly credits included in the plans, for higher-cost models, usage is balanced through 5-hour and 7-day rolling windows.
| Plan | 5-Hour Window | 7-Day Window | Relative Ratio |
|---|---|---|---|
| Starter | ~ 2,720 requests1.2M credits | ~ 9,090 requests4M credits | 1× |
| Plus | ~ 5,450 requests2.4M credits | ~ 18,180 requests8M credits | 2× |
| Pro | ~ 16,360 requests7.2M credits | ~ 54,540 requests24M credits | 6× |
Request counts are estimated at GPT-6 Luna prices, using 2,000 uncached input tokens, 12,000 cached input tokens and 240 output tokens per request. GPT-6 Luna is used only as a pricing reference and has no window limit.
Frequently Asked Questions
We accept credit/debit cards (some UnionPay cards may not be supported), as well as Alipay, Apple Pay, Google Pay, Link, Amazon Pay, and Cash App Pay.
If you haven't used any credits, you can request a full refund within 30 days of subscribing. If you've only used a little and it isn't working out, reach out to us—we review these case by case and try to make it right. Full details are in the refund policy linked at the bottom of the page.
You can upgrade or downgrade your plan at any time from your account settings. When upgrading, you'll be charged the new price. Downgrades take effect in the next billing cycle.
You can cancel your subscription at any time from your customer portal. After cancellation, you will continue to enjoy the current subscription service until the end of the billing cycle, and you will not be charged thereafter.