Published token rates. OpenAI-compatible.
Drop in the endpoint, pay per token. Compare us directly against any provider.
Pricing
One subscription. Three services. Priced to make the switch obvious.
Drop in the endpoint, pay per token. Compare us directly against any provider.
Text inference, tool calls, and image generation in one seat. Commit quarterly and pay less than monthly billing.
Reserved capacity, security controls, and custom retention belong on contract. Not hidden inside self-serve pricing.
Pricing
Every plan includes text inference. Add tool calls and image generation as you grow.
$0.25 / 1M input
$1.00 / 1M output · pay only for what you use$29 every 90 days
Under $10/mo · save 20% vs monthly$99 every 90 days
$33/mo · save 15% vs monthly$199 every 90 days
$66/mo · save 16% vs monthly$45 per seat / 90 days
Five-seat minimum · $15/seat/mo$225 per seat / 90 days
Five-seat minimum · $75/seat/moPublished rates match your invoice. No platform fees, no per-seat markups on tokens.
Weekly limits reset on a rolling window, not a calendar month. Use your allocation whenever you need it.
Need reserved throughput, custom retention, or security controls? Talk to us.
Use code BETACLAW after signup for +20% extra rate limits on any paid plan.
Earn $5 credit for every paying friend you invite. Save your Claude tokens for deep work — tell them to try tk_ for everyday tasks.
tk_
Run your prompt against the benchmark wall, compare the output, and switch when the evidence is obvious.