Pricing

Let us build it
together.

No fixed tiers, no surprise bills. We work with each studio to size a licensing deal around your game - embedded on-device or hosted in the cloud. Same model, same characters, same quality.

Get in touch to work together

We tailor every licensing deal to the studio. Tell us about your game - your platform, your scale, and whether you want the model embedded on-device or hosted in the cloud - and we will put together the right package for you.

Or email us directly at [email protected]

Embedded or Cloud?

Here is the quick decision tree - we will help you pick when we talk.

Choose Embedded if...

  • Your game ships on PC or console
  • You want zero latency (20ms)
  • Players should work offline
  • You prefer a one-time license over a subscription
  • You want zero per-player cost at scale

Choose Cloud if...

  • Your game is mobile, web, or cloud-streamed
  • You cannot bundle a 1.9 GB model
  • You want auto-updating model versions
  • You want to prototype before committing
  • You are building a live-service or MMO

Start on Cloud, ship on Embedded. Many studios prototype with Cloud, then switch to Embedded when they are ready to ship. Your characters, training, and personality definitions transfer with zero rework.

No per-install fees

We learned from Unity. You will never be charged based on installs, downloads, or player counts.

No retroactive changes

Your license terms are locked for the lifetime of your title. Future changes only apply to new licenses.

Your data, always

Export characters, training data, and conversation history at any time. Embedded games work forever - even if we vanish.

FAQ

Common questions

Embedded ships the model inside your game binary - it runs locally on the player's machine with 20ms latency, works offline, and has zero per-call cost. Cloud is the same model hosted on our GPUs - you call our API, we handle inference. Same model, same quality, different delivery.

Yes. The model is the identical GGUF in both tiers. Your characters, personality definitions, and training transfer directly. Swap the SDK endpoint from our API to localhost and you are done. Many studios prototype on Cloud and ship on Embedded.

Reach out at [email protected] and tell us about your game - the engine, the platform, how many characters you expect, and whether you are leaning embedded or cloud. We will scope the right setup with you. You can also try the model live in the playground first, no signup required.

The model runs on consumer GPUs (RTX 3060 and above) and Apple Silicon (M1 and above). 1.9 GB VRAM required. Most gaming PCs built in the last 3 years handle it without issues.

No. Your license terms are locked for the lifetime of your title. Whatever we agree on is fixed per-title - any future changes apply only to new agreements.

Embedded games work forever - the model runs locally with no server dependency. For Cloud customers, all character data, training data, and conversation history can be exported at any time. Your data is always yours.

Ready to give your NPCs a voice?

Try it live in the playground, then get in touch to work together.