One key.
Open-weight AI models — DeepSeek, Qwen, Kimi, Llama and more — through a single OpenAI-compatible key. One Points balance, pay per token.
Every model is one string away
Keep your code. Change the model name — the key, the balance and the bill stay the same.
client.chat.completions.create( model="deepseek/deepseek-v3.1", messages=[{"role": "user", "content": "…"}], )
Explain what a reverse proxy does in one sentence.
A reverse proxy sits in front of your servers, takes requests from clients and forwards each one to the right backend — so clients only ever talk to one address.
Starting balance
100 ptsWorks with: Cursor, Cline, LangChain, n8n, Roo Code, Kilo Code, Open WebUI, LlamaIndex, AI SDK, Dify, Langflow, Goose, Zed, JetBrains AI, Obsidian
Cursor
LangChain
Roo Code
Open WebUI
AI SDK
Langflow
Zed
Obsidian
Cline
n8n
Kilo Code
LlamaIndex
Dify
Goose
JetBrains AIOpenAI-compatible
Same SDK, same request shape. Two lines change: base_url and the key.
A console you'll actually open
Keys with budgets, a chat to try any model, and spend you can read at a glance.
Try any model.
A built-in chat, billed to the same balance.
Requests
last 14 days
Spend this month
By modelProduction key
sk-taldermont-••••7f3a
Allowed models
Balance
2,380 ptsKey limit
of 10,000 pts this month
On track
Field guide
6 files
Pay only for tokens
No subscriptions. Buy Taldermont Points at a fixed rate of 100 Points = USD 1, and each request deducts Points at the model's listed price. Unused paid Points can be refunded on request.
- Price per 1M tokens and per typical answer
- A receipt for every purchase
- Spend limits on every key
Questions people actually ask
So what am I actually paying for?
For tokens — the text a model reads and writes. You buy Taldermont Points once (100 Points = USD 1), and every request deducts Points at the price listed next to the model. No subscription, no minimum monthly spend, no surprise overage.
Will I have to rewrite my code?
No. If your app talks to the OpenAI API, change two lines — base_url and the key — and you're done. The same goes for Cursor, Cline, n8n, LangChain and anything else that accepts a custom endpoint.
Which countries and payment methods do you support?
We work with customers in most countries — the list of places where the service isn't available is in our Terms of Service. You pay by card in seven currencies, and companies can ask for an invoice for a card purchase. Points are bought at a fixed rate of 100 Points = USD 1, so the price of a request doesn't jump around with exchange rates.
What if a model goes down on the provider's side?
If the model has a backup source, the request is quietly retried there. If there's no healthy source left, you get a clear error and we don't charge you for it.
Do you read or keep my prompts?
We don't use them to train models, and by default we store only metadata: which model, how many tokens, what it cost. Your prompt goes to the model provider (currently DeepInfra) to generate the answer, and the provider handles it under its own terms. You can turn on content logging for debugging — texts are then kept for 30 days.
Can I get my money back?
Yes. Unused paid Points are refunded on request at their original purchase value. If you're in the EU or UK, you also have a 14-day right of withdrawal — the details are in our refund policy.
How do I try it before paying?
Sign up and you get 100 free Points straight away — enough for a few hundred answers from a budget model. No card needed.
Why not rent GPUs or sign up with a hosting provider directly?
You can. We're useful when you want many open models without several accounts, cards and invoices; when you need a budget per key; or when you want one place to compare prices and switch models by changing one line.