Basic package
Add $3 credit
3
USD credit
Expiry: one year
Pricing
Pay-as-you-go
Choose one of the top-up packages; the cost is deducted from your credit only when you use each model. Each model's approximate rate is in the tables below on this page.
Quick jump
Top-up packages
Buy credit, use as much as you need.
Each package adds a specific amount of credit to your account. After top-up, that credit is spent according to each model's rate with each use.
All users have access to every capability and available model.
Packages differ only in the amount of credit you buy; there's no restriction on tool or model type between packages. Actual usage depends on your request volume.
With the same credit you can access these models and the rest available in the app
Best for light users
Add $8 credit
8
USD credit
Expiry: one year
Economical everyday use
SpecialAdd $20 credit
20
USD credit
Expiry: one year
Professional everyday use
Add $45 credit
45
USD credit
Expiry: one year
Pay-as-you-go
Rate list for each model
Each AI model has a different rate. This rate is deducted from your wallet when you use ChatQT. In the chat tool, each chat's cost is shown at the top of the page, and in the other tools you can see a model's rate by selecting it. In the table below you can review the list of models and rates.
Chat & reasoning
Unit for all models in this section: per 1000 characters of prompt or response text
| Model | Unit | Your prompt (USD) | Model response (USD) |
|---|---|---|---|
| Claude Opus 5 | Per 1000 characters of prompt or response text | — | — |
| Claude Fable 5 | Per 1000 characters of prompt or response text | — | — |
| GPT 5.6 Sol | Per 1000 characters of prompt or response text | — | — |
| GPT 5.6 Terra | Per 1000 characters of prompt or response text | — | — |
| GPT 5.6 Luna | Per 1000 characters of prompt or response text | — | — |
| GPT 5.5 | Per 1000 characters of prompt or response text | — | — |
| Claude Opus 4.8 | Per 1000 characters of prompt or response text | — | — |
| Claude Sonnet 4.6 | Per 1000 characters of prompt or response text | — | — |
| Gemini 3.5 Flash | Per 1000 characters of prompt or response text | — | — |
| Gemini 3.1 Pro | Per 1000 characters of prompt or response text | — | — |
| Grok 4.5 | Per 1000 characters of prompt or response text | — | — |
| Grok 4.1 Fast | Per 1000 characters of prompt or response text | — | — |
| Kimi K 2.5 | Per 1000 characters of prompt or response text | — | — |
| DeepSeek V4 Pro | Per 1000 characters of prompt or response text | — | — |
| DeepSeek V4 Flash | Per 1000 characters of prompt or response text | — | — |
| DeepSeek V3.2 | Per 1000 characters of prompt or response text | — | — |
| Qwen 3.7 Max | Per 1000 characters of prompt or response text | — | — |
| GLM 5.2 | Per 1000 characters of prompt or response text | — | — |
| GLM 5 | Per 1000 characters of prompt or response text | — | — |
Image
| Model | Unit | Approx. USD |
|---|---|---|
| Nano Banana Pro | One image generation (approx. 1K) | — |
| Nano Banana | One image generation (approx. 1K) | — |
| GPT Image 2 | One image generation (approx. 1K) | — |
| GPT Image 1.5 | One image generation (approx. 1K) | — |
| FLUX.2 Pro | One image generation (approx. 1K) | — |
| Imagen 4 Ultra | One image generation (approx. 1K) | — |
| Grok Image | One image generation (approx. 1K) | — |
| Seedream 4.5 | One image generation (approx. 1K) | — |
| SD Ultra | One image generation (approx. 1K) | — |
Video
| Model | Unit | Approx. USD |
|---|---|---|
| Sora 2 | Per second at 720p | — |
| Sora 2 Pro | Per second at 720p | — |
| Gemini Omni Flash | Per second of video | — |
| Veo 3.1 | Per second at 720p | — |
| Veo 3.1 Fast | Per second at 720p | — |
| Seedance 2.0 | Per second of video | — |
| Happy Horse 1.1 | Per second of video | — |
| Kling 3.0 Pro | Per second at 1080p | — |
| Kling 2.6 Pro | Per second at 1080p | — |
| Minimax | Per second at 720p | — |
| Wan 2.2 | Per video | — |
| Hunyuan Video | Per video | — |
Audio and speech
| Model | Unit | Approx. USD |
|---|---|---|
| OpenAI TTS 1 HD | Per 1K characters | — |
| ElevenLabs Multilingual | Per 1K characters | — |
| Google TTS Studio | Per 1K characters | — |
| Whisper | Per minute of audio | — |
| QTstt | Per minute of audio | — |
Selection guide
Which top-up package is right for you?
Every package gives the same software access; only the starting credit amount changes.
| If… | Our suggestion |
|---|---|
| Don't trust ChatQT yet? | The $3 package to test |
| You're just starting out or use cheaper models like GPT Mini and Grok. | The $8 package; enough for about three weeks to a month |
| You use it daily and across a variety of tools. | The $20 package |
| You seriously try several models and tools each week, or have a small team and higher usage. | The $45 package |
| Your volume or contractual needs exceed the packages. | Enterprise — SLA, official invoicing and custom capacity |
Comparison
ChatQT or separate subscriptions for ChatGPT and Claude?
ChatQT is several times more cost-effective. On ChatQT you have one shared credit to use all the AI models, with no need for separate subscriptions. Each model has a different rate; from the Usage rates section, it's viewable.
One credit, all kinds of AI
Without multiple back-to-back subscriptions for each maker.
Switch the top model
Every few weeks a company releases a new model that's better than the rest. Switch with one click.
No daily usage cap and …
Multiple models without managing separate daily quotas.
All kinds of models, in one app
No need for separate subscriptions for each maker.
Pay-as-you-go
Cost per use in Usage rates.
chat, image, video and audio
all in the same app and with the same credit.
pay-as-you-go
Buy credit once and spend only on what you use.
Enterprise
Teams and long-term contracts.
The same pay-as-you-go product with credit top-ups; for high-volume organizations: SSO, seat management, SLA, official invoicing and dedicated onboarding.
- ✓ The same features and usage models with custom capacity and billing
- ✓ Contract and volume discount
- ✓ 24/7 support and team training
What does ChatQT's pay-as-you-go mean?
First, from among the Top-up packages (or the payment section in the app) you buy credit; then you have access to every capability and available model, and the cost is deducted from your credit only for actual usage — prompt and response text, each image generation, each second of video, and so on. Before sending, the app usually shows the approximate cost; the tables on this page are for comparing models.
Do I really get the same ChatGPT or Claude model?
The model name you select in the app is the same model family from the same maker; behavior and quality match that name and model tier. ChatQT is a unified usage and payment layer; the table on Usage rates this same page shows the model names and the approximate cost per unit. To compare with a separate ChatGPT or Claude subscription, see the Compared to a subscription section.
Why not just buy several separate official subscriptions?
Every few weeks one of the flagship models takes the lead; with one ChatQT credit you switch models in the app without buying a parallel subscription for each brand. For professional work you usually need several models, and with separate subscriptions you often hit each product's daily quota. A summary is in the Compared to a subscription section.
What's the difference between the top-up packages?
Only amount of credit of the package differ. You have equal access to all tools and models in the service; the larger the package, the more easily you can cover heavier usage.
How do I top up credit and how does expiry work?
From the payment section in the app, you choose one of this page's packages or coordinate with support; after payment, credit is added to your account. For each package, the credit validity is shown on the package card on this page; the balance is valid until then or until it runs out, unless stated otherwise in the app.
Which payment methods are supported?
You top up credit in the app with your selected payment method. For enterprise contracts, an official invoice is issued.
If I regret the purchase, is a refund possible?
Yes, within 7 days of purchase, if you're not satisfied, the amount paid is fully refunded.
Does my data stay private?
Your data is never used to train our models and is stored encrypted per the privacy policy.
Do you offer discounts for students or startups?
Yes, we have special programs for students, faculty and early-stage startups. Contact us for more information.