Pricing
Pick the plan that fits.
Run it free on your own machine, or let us host the model in a confidential-compute enclave so you don’t need a GPU at home. You can switch tiers anytime.
Self-Host
0€
You host it. Everything runs on your machine.
- Bridge daemon included — macOS, Windows, Linux
- Chat, image, video, code — fully local, offline
- Every open-weight model, uncensored, on your GPU
- BYO runtimes (Ollama, LM Studio, ComfyUI, MLX)
- No account, no telemetry, no cloud
- Open-source (AGPL-3.0)
Hosted
19€ / mo
We host it. The same models, no GPU needed.
- Everything — text, image & video in the cloud
- Every uncensored open-weight model
- TEE-attested: privacy verifiable on every response
- Unlimited text (fair use)
- Media credits ≈ 160 images or 10 video clips / month
- Works in any browser, nothing to install
Most popular
Hosted Pro
49€ / mo
More of everything. The sweet spot.
- Everything in Hosted, higher quota
- Media credits ≈ 560 images or 35 video clips / month
- Higher resolution, longer video clips
- Multiple parallel jobs, priority queue
- Workflow import + credit banking
Hosted Max
99€ / mo
Maximum quota for power users.
- Everything in Hosted Pro, maximum quota
- Media credits ≈ 1600 images or 100 video clips / month
- Highest priority, maximum parallelism
- Longest videos, full API limits
- Extended credit banking
Payments processed securely by Stripe. 14-day refund, no questions asked. Read the terms.