Your minions run on tokens. Tokens are what they feed on to stay alive and keep working — a hungry minion is a happy minion. Pick a plan, keep them fed, and they never clock off. One bundle covers every model.
Starter
S$55/mo
23.4M tokens · 1 minion
1 dedicated minion VM
~S$10 of credits monthly
Web, Telegram & WhatsApp
Top up anytime (top-ups never expire)
Minion's email included
Minion workspace
Daily backups
BYOK (bring your own key)
SOONSquad
POPULARS$125/mo
93.6M tokens · 2 minions
2 dedicated minion VMs
~S$40 of credits monthly
Web, Telegram & WhatsApp
Priority support
Minion's email included
Minion workspace
Daily backups
BYOK (bring your own key)
SOONEnterprise
Custom
Custom tokens · unlimited minions
Customized & self-hosted models
Private VM cluster
SLA + dedicated support
Volume credit pricing
Minion's email included
Daily backups
Minion workspace
BYOK (bring your own key)
SOONYour tokens work across every model. Lighter models stretch your balance further; premium models draw more per message. Focus on what matters — pick the model that fits and switch anytime.
DeepSeek v4 Flash
Recommended for most use cases
1x
base rate
GPT-5.4
Premium quality
8.9x
tokens used
GPT-5.4 mini
Fast & affordable GPT
8.9x
tokens used
Kimi K2.6
Long-context, security analysis
10.7x
tokens used
Kimi K2.5
Long-context, budget option
10.7x
tokens used
DeepSeek v4 Pro
Best for complex reasoning
12.4x
tokens used
GPT-5.5
Top-tier, highest cost
12.5x
tokens used
GLM 5.1
via MuleRouter
15.7x
tokens used
Rates are multiples of the base model — a 14× model uses 14 of your tokens for every 1 the base model would. Repeated context is cached and billed at a fraction, so follow-up messages cost less.
Tokens are the unit your minion spends when it reads and writes. One bundle covers every model we route to, so you never pick a model or juggle separate bills.
Four minutes to set up. One conversation to start working.
