Without cache
- 50,000 new input tokens$0.10 per 1M
- 2,000 output tokens$0.50 per 1M
60,000 tokens
Usage, explained
How Social Bellow tokens work
One usage unit, priced from what the AI model costs. Pick a tab to see where the tokens go.
Your tokens at a glance
Chats, captions, design edits and images all use Social Bellow tokens. Each request costs what the AI model costs, so cheaper models use fewer tokens.
- One unit for every model
- Monthly allowance on paid plans
- See your usage in Settings
Social Bellow tokensPro plan
123,548,000 left this month
1,452,000 used125,000,000 a month
- 12 × short chats54,000
- 2 × long chats48,000
- 1 × premium chat90,000
- 3 × images1,260,000
1,000,000tokens = $0.10 of model cost, the base rate
1token for each input token on our base model
4,500tokens for a short chat on our base model
125,000,000tokens in a Pro month, for $19.99
How a charge is made
From prompt to Social Bellow tokens
The same steps happen on every request, on every model.
- Your promptThe words you send and the reply the model writes.2,000 in, 500 out
- Tokens countedInput and output are counted at the model’s rates.2,500 tokens
- USD model costWhat the provider charges for this request.$0.00045
- ÷ $0.10 per 1MDivide the cost by the base rate.× 10,000,000
- Tokens debitedThe result is the request’s Social Bellow token charge.4,500 tokens
- Allowance firstYour monthly allowance is used first, then any extra tokens.Then extra tokens
Model choice
Bigger models use more tokens
The same prompt, 2,000 tokens in and 500 out, on four models.
Bars share one scale. The multiple compares each model with the cheapest one. Rates are list prices per 1M tokens.
Cache savings
Repeated context costs far less
Cached input on our base model is $0.01 per 1M tokens, a tenth of the $0.10 base rate. The same long chat, billed two ways:
60%fewer tokens
With cache
- 10,000 new input tokens$0.10 per 1M
- 40,000 read from cache$0.01 per 1M
- 2,000 output tokens$0.50 per 1M
24,000 tokens
A follow-up in the same chat often costs less than the first message. For a new task, start a new chat.
What 1,000,000 tokens gets you
About this much work for one million tokens
One million tokens is $0.10 of model cost at the base rate. Here is roughly what that buys, using the worked examples on this page.
Counts are rounded down to whole requests.
222short chats on our base model
41long chats, with cache
11premium chats on GPT-6.1 Sol
2images on GPT Image 2.5 Sunburst
125Mtokens a month
A Pro month
Where 125 million tokens might go
An example month, not a fixed split. Each part is the request count times its token cost.
- Short chats · 8,000 requests36,000,00028.8%
- Long chats · 1,000 requests24,000,00019.2%
- Premium chats · 200 requests18,000,00014.4%
- Images · 100 requests42,000,00033.6%
- Left over5,000,0004.0%
Try it
Estimate one message
Pick a model and the length of your message and reply.
Model cost: $0.10 per 1M input tokens, $0.50 per 1M output tokens
Estimated charge
4,500 tokens
$0.00045 of model cost ÷ $0.10 per 1MAbout 27,777 messages like this fit in one Pro month.
Estimates use list rates and leave out cache, images and any extra tool calls.
Plans
Basic and Pro
Both plans include a monthly allowance of Social Bellow tokens that resets every month.
Basic
A bigger monthly allowance for regular posting.
$9.99/month
$9.99/month, billed monthly
Get Basic- 50 million Social Bellow tokens per month
- Image generation
- Web search
- Canva & Adobe Express add-ons
- Unlimited saved prompts
Pro
The largest allowance, plus the Pro-only models.
$19.99/month
$19.99/month, billed monthly
Get Pro- 125 million Social Bellow tokens per month
- Earlier frontier model access
- More Image generation
- Web search
- Canva & Adobe Express add-ons
- Unlimited saved prompts
On our base model, Basic covers about 11,000 short chats a month and Pro about 27,000.
Not ready to pay? A free account gets a few uses on our base model to try it. Start free
Order of use
Your monthly allowance goes first
Each request is paid from your plan allowance. Extra tokens are only used once the allowance runs out.
Step 1
Monthly plan allowance
Used first. It resets every month.
Step 2
Extra tokens
Used after the monthly allowance is used up.
See your usage in Settings. Open the Tokens tab to see the tokens you have used this month and how many are left.
No. A model token is a piece of text. A Social Bellow token is our usage unit, set by what the AI model costs for your request. That is why the same amount of text can use different numbers of Social Bellow tokens on different models.
Ready to try Social Bellow tokens?
Start free on our base model. Upgrade to Basic or Pro when you need a bigger monthly allowance.