Gateway livePrices
(00) underlist

Every model,
under list.

One API key for Claude, GPT, Gemini and 460+ more. Credits bought under list price, spent token by token.

Net, $100 top-up · demo−23.5%
Claude Opus 5.5$20.00$15.30Claude Sonnet 5$10.00$7.65GPT-6 Sol$10.00$7.65GPT-6 Luna$0.50$0.38Gemini 3.8 Flash$3.75$2.87Grok 4.7$6.00$4.59DeepSeek V4.1 Flash$1.00$0.77Qwen3.8 Max (0902)$6.00$4.59Kimi K3$12.30$9.41Mistral Medium 3.5$7.50$5.74Claude Opus 5.5$20.00$15.30Claude Sonnet 5$10.00$7.65GPT-6 Sol$10.00$7.65GPT-6 Luna$0.50$0.38Gemini 3.8 Flash$3.75$2.87Grok 4.7$6.00$4.59DeepSeek V4.1 Flash$1.00$0.77Qwen3.8 Max (0902)$6.00$4.59Kimi K3$12.30$9.41Mistral Medium 3.5$7.50$5.74per 1M output · demo book
(01)Top up

Buy credits under list.

Pick an amount. The quote fills from the cheapest offers on the book and shows every line before you pay.

1 credit = $1 of usage at the model's list price.

Top up credits

You pay…

Net of the 5% fee, vs list price. Card checkout by Whop; processing shown before you pay.

From $5 to $2,000
per card purchase
5% platform fee
on face value, on the receipt
Card, Apple Pay, Google Pay
through our checkout partner
On your balance
as soon as the payment clears
(02)Price check

Same answer. Smaller bill.

Pick a prompt and a model. The answer is a saved one, replayed; the list price is the model's live price on OpenRouter.

The demo replays a stored answer: no model is called.

Your prompt

Why does Postgres say column "user_id" is ambiguous?

Try another prompt

Answer
demo · replayed answer

anthropic/claude-sonnet-5

ready

Before · 1M output

$10.00

With underlist

$7.65
You save$2.35 −23.5%
list price live · discount demo

The discount is a book.

Credits are listed at set discounts under list price. Your order takes the cheapest first; whatever the book can't fill is sold at list price, never above.

How the book works →

Offers
Recent buys
settled purchases

Loading…

(04)Models

467 models. Priced live.

List prices load live from OpenRouter. The underlist column applies the net rate of a $100 top-up on the demo book, card fee included.

Called by their usual ids. The model you name is the model that runs.

Model prices
live list pricesdemo discount
Modelunderlist · 1M out
Claude Haiku 5.5
anthropic/claude-haiku-5.5
$0.38−23.5%
Nano Banana 2.1
google/gemini-nano-banana-2.1
$5.74−23.5%
Mistral Large 4
mistralai/mistral-large-4-0
$1.60−23.5%
GPT-6.1 Sol Pro
openai/gpt-6.1-sol-pro
$7.65−23.5%
GPT-6.1 Sol
openai/gpt-6.1-sol
$7.65−23.5%
Claude Sonnet 5.5
anthropic/claude-sonnet-5.5
$7.65−23.5%
Qwen3.8 Max Prime
qwen/qwen3.8-max-prime
$9.18−23.5%
GPT-6 Luna Pro
openai/gpt-6-luna-pro
$0.38−23.5%
GPT-6 Luna
openai/gpt-6-luna
$0.38−23.5%
GPT-6 Sol Pro
openai/gpt-6-sol-pro
$7.65−23.5%
GPT-6 Sol
openai/gpt-6-sol
$7.65−23.5%
Claude Opus 5.5
anthropic/claude-opus-5.5
$15.30−23.5%
12 of 80 shown · 467 models available
(05)Method

Four steps. No contract.

  1. 01

    Top up

    From $5 by card. Your order fills from the cheapest offers on the book, and the receipt shows every line first.

  2. 02

    Take one key

    One key for every model. Rotate or revoke it whenever you like: your balance never moves.

  3. 03

    Swap two lines

    The base URL and the key. OpenAI and OpenRouter SDKs, streaming and tool calls carry on as they are.

  4. 04

    Pay per token

    Every call is charged at the model's real cost, from your credits. No plan, no seat, no monthly fee.

Two lines. That's it.

Point your client at underlist and swap the key. Model names, streaming and tool calls keep working as they are.

Read the API docs →

Base URL

https://api.undrlst.com/v1

API key example

sk-smaaart-7f3a9c2e1b…

.env+2 lines
# any OpenAI-compatible client reads these
OPENAI_BASE_URL=https://openrouter.ai/api/v1
OPENAI_API_KEY=sk-or-v1-…
OPENAI_BASE_URL=https://api.undrlst.com/v1
OPENAI_API_KEY=sk-smaaart-…

Every token, on the record.

Your balance, spend by day and by model, and the latency our gateway adds. Shown here with a sample account.

Open your dashboard →

undrlst.com/account
sample account
Balance
$142.57
Spent, last 7 days
$112.00
Requests, last 7 days
4,927
Added latency (median)
14 ms

Spend by day

Top models

  • anthropic/claude-sonnet-5$48.20

    892 requests · 43%

  • openai/gpt-6-sol$36.10

    645 requests · 32%

  • google/gemini-3.8-flash$19.40

    1,234 requests · 17%

  • deepseek/deepseek-v4.1-flash$8.30

    2,156 requests · 7%

Asked. Answered.

The docs cover the API in detail →

Search underlist

Search pages, models, docs and questions