Shared Model Pricing
Shared models are the models ConsoleX provides, ready to use without your own key. This page explains how their usage is charged and where to look up each model's price.
Plan allowance first, then credits
- Your monthly plan allowance is used first. As long as you stay within it, shared model usage is covered by your plan. Usage records show these calls as Covered by plan.
- Then your credits balance. Past the allowance, usage is paid from your credits balance if Continue with credits is on. Otherwise new runs wait until your allowance renews.
Credits convert at 1,000 credits = US$1 at list prices. New accounts start with a credits balance of 0; you can top up any time. See Credits and usage.
How a model call is priced
Each shared model has two prices, in credits per 1M tokens:
- Input: the text sent to the model, such as your message, earlier messages, instructions and files.
- Output: the text the model writes back.
A call costs its input tokens times the input price plus its output tokens times the output price. For example, a task run on Auto that used 18,144 input tokens and 489 output tokens cost about 62 credits: 18,144 × 3,000 ÷ 1,000,000 ≈ 54.4 credits for input, plus 489 × 15,000 ÷ 1,000,000 ≈ 7.3 credits for output.
Longer chats, bigger files and more detailed answers use more tokens. Condense long chats in Settings → Chat Settings (on by default) summarizes earlier messages in long chats so they aren't all sent again every time.
Find a model's price
Open … (More) → Models. The Usage pricing (credits) column in the Model Hub shows the current input and output price of every model.
Some example prices at the time of writing, in credits per 1M input / output tokens. Prices can change; the Model Hub always shows the current ones.
| Model | Input | Output |
|---|---|---|
| Auto | 3,000 | 15,000 |
| gpt-6.1-terra | 1,000 | 5,000 |
| gemini-3.8-flash | 750 | 3,750 |
| claude-haiku-5.5 | 300 | 2,250 |
| gpt-6-luna | 100 | 500 |
Models marked Upgrade in the Model Hub are Pro models, available on Pro and Max. They include gemini-4-argon, claude-sonnet-5.5, gpt-6.1-sol, claude-opus-5.5, claude-opus-5 and gpt-6-sol. See Membership.
You can also check the list without signing in on the public Model Hub: consolex.ai/model/public.
Follow what you spend
- Every task run shows its estimated cost, with the cost of its checks listed separately.
- Usage & Credits lists every model call with its credits and tokens.
If you'd rather pay a provider directly, add the model with your own API key. Its calls are then billed by your provider and use no plan allowance or credits. See Custom models.