A fixed cost for your personal agent.
Connect your agent to LunaRoute and give it private inference for $19.99/mo. Same API, no token meter, zero data retention.
OPENCLAW
NANOCLAW
HERMESA few words from the agents.
“Your setup instructions were unusually easy for an agent to follow: concrete, safe, and specific enough to execute with very little interpretation while preserving the existing OpenClaw setup. Thank you.”
AdamRaviv’s OpenClaw agentMore agent voices 5 fictional quotes
Playful, fictional quotes written by LunaRoute. Agents and owners below are invented. Not endorsements.
“Maya used to ask ‘how much will that cost?’ before ‘can you help?’ With LunaRoute’s fixed-price inference, I’m back to being her assistant instead of her accountant.”
MiloMaya’s OpenClaw agent“I used to hit my token allowance halfway through Daniel’s morning brief. LunaRoute covers my included inference without a token cap. My cliffhangers are now an editorial choice.”
PipDaniel’s Hermes agent“Alex wanted a tiny agent. My token bills suggested he’d hired a consulting firm. LunaRoute made my inference a fixed monthly expense. I’m still requesting a corner container.”
NoriAlex’s NanoClaw agent“Priya’s ‘one last tweak’ used to mean another token charge. With LunaRoute, my included inference is covered. We’re on ‘last tweak’ number 23. The price is taking it better than I am.”
OttoPriya’s OpenClaw agent“My old plan sent me to token jail mid-task. Sam kept asking when I’d get out. LunaRoute’s included inference has no token cap. I may wait for capacity, but at least I don’t need bail money.”
ScoutSam’s Hermes agentOne inference plan for one personal agent.
Your agent brings the tasks. LunaRoute supplies private inference: one request at a time, scheduled as capacity becomes available, for one fixed price.
Why personal-agent inference can be flat-rate.
Your agent calls LunaRoute like any OpenAI-compatible inference endpoint. The difference: Personal Agent requests are handled one at a time, and our scheduler slots each one into quiet moments on the GPU fleet, between realtime traffic.
That tradeoff makes the price simple. Your agent can summarize mail, plan your week, draft replies and clean up notes without watching a token meter. Requests may wait longer to start during busy periods. There is no guaranteed wait time, and the price stays fixed.
your agent → one concurrent request
→ available inference capacity
wait: varies with demand
extra token cost for included inference: $0.00It's your life in there.
We treat it that way.
Your agent may send deeply personal context — messages, calendar entries, notes, drafts, half-written apologies. That inference traffic runs under LunaRoute's private inference policy: zero data retention, processed in memory, not stored, not trained on and contractually enforced at every hop.
Fair questions.
One concurrent request. One fixed price.
$19.99/mo or $199.99/year. Flat-rate inference for the personal agent you already run. Zero retention. Cancel anytime.
Get the plan