Cost-efficient private AI inference
Darkbloom routes encrypted requests to hardware-verified Apple Silicon providers, delivering comparable model performance at about 50% lower cost than typical API providers. Prompts stay hidden from operators, and Mac owners earn from compute they already own.
Private inference without a new SDK
Change the base URL and keep your existing OpenAI client. Requests are encrypted before they leave your app and routed to verified Apple Silicon providers.
Open Console ↗Turn idle Apple Silicon into earnings
Run a provider on hardware you already own. Darkbloom matches your Mac with inference demand, and operators keep 100% of inference revenue during the public alpha.
Start Earning ↗Capacity is bought, rented, repackaged, and metered before it reaches an API call. Each layer adds margin. Darkbloom routes demand to idle Apple Silicon instead, where the hardware is already paid for and the marginal cost is mostly electricity.
→
→
API providers
→
End users
Darkbloom turns that idle capacity into a private inference market.
Developers get lower prices without changing SDKs. Mac owners earn from machines they already own. The coordinator matches demand to providers, but prompts stay encrypted and hidden from the operator.
Prompts can contain customer conversations, internal plans, source code, and other sensitive context. A marketplace promise is not enough when inference runs on hardware you do not own.
Darkbloom is designed around a stricter guarantee: the coordinator can route requests, the provider can serve them, but neither should get a usable view of the prompt.
Private inference requires privacy that can be verified, not just promised.
Operator-blind by design
Darkbloom removes the practical software paths an operator could use to observe inference data. Four layers work together, each independently verifiable.
Encrypted end-to-end
Requests are encrypted before transmission. The coordinator routes ciphertext, and only the matched provider's hardware-bound key can decrypt the request.
Hardware-verified
Each provider uses a key generated inside Apple's tamper-resistant secure hardware. The attestation chain traces back to Apple's root certificate authority.
Hardened runtime
The inference process is locked down at the OS level. Debugger attachment and memory inspection are blocked so the operator cannot inspect a running request.
Traceable to hardware
Responses are signed by the specific machine that produced them. The attestation chain is public, so users can verify the hardware behind the result.
The operator contributes compute, not visibility.
Your prompt is encrypted before it leaves your app. The coordinator routes traffic it cannot read. The provider serves the request inside a hardened process the operator cannot inspect.
Read the paper ↗OpenAI-compatible API
Keep your SDK, request shape, and streaming code. Point the client at Darkbloom and start routing private inference.
from openai import OpenAI
client = OpenAI(
base_url="https://api.darkbloom.dev/v1",
api_key="your-api-key"
)
response = client.chat.completions.create(
model="gemma-4-26b",
messages=[{"role": "user", "content": "Hello!"}],
stream=True
)
for chunk in response:
print(chunk.choices[0].delta.content, end="")50% lower cost, comparable performance
Idle Apple Silicon keeps the cost structure simple. Pay per token with no subscription or minimum, with selected model prices set around 50% below typical API-provider rates for comparable models.
| Model | Input | Output | Typical API | vs typical API |
|---|---|---|---|---|
| Gemma 4 26BMoE · 128K context | $0.03 | $0.165 | $0.33 | 50% lower |
| GPT-OSS 20BMoE · 128K context | $0.015 | $0.07 | $0.14 | 50% lower |
Prices per million tokens. Typical API means published list rates for comparable models from major API providers.
Earn from your Mac
Install the provider, choose when your Mac is available, and earn from inference jobs matched by the network. During the public alpha, operators keep 100% of inference revenue.
Install via Terminal
Downloads the provider binary and configures a background launchd service.
$ curl -fsSL https://api.darkbloom.dev/install.sh | bashEarnings estimate
What your Apple Silicon Mac could earn serving inference on the Darkbloom network.
Estimated annual earnings
… /yr
This is a projection, not a promise. Usage earnings assume healthy, sustained network demand — live demand fluctuates and can run below this. The base reward requires an attested, healthy machine that stays online ≥90% of each settlement period, and is paid from a fixed monthly pool — an earnings floor while eligible, not a guarantee.
Live network catalog. Your earnings estimate uses the best-earning model automatically.
Base rewards (earnings floor)▾
On top of what you earn from real inference, attested machines accrue a monthly base reward for staying online. It is paid in 5-minute prorated increments, while real usage remains the upside on top.
| Unified memory | Per month | Per year |
|---|---|---|
| 512GB | $40 | $480 |
| 192GB | $30 | $360 |
| 128GB | $26 | $312 |
| 96GB | $22 | $264 |
| 64GB | $18 | $216 |
| 48GB | $16 | $192 |
| 32GB | $12 | $144 |
| 24GB | $10 | $120 |
| Under 24GB | — | — |
- Rewards settle every 5 minutes and require staying online ≥90% of that period.
- Usage earnings are paid on top of the base reward — you keep 100% of both.
- Base rewards go to attested, online, healthy machines up to a fixed monthly budget, prorated per settlement period; not a guarantee.
How we estimate this▾
The top of the range assumes healthy, sustained demand: your Mac serves requests 60% of the time it is online, averaging 2.5 concurrent requests while active — not a saturated network (the engine can batch more at peak). Live demand fluctuates and can run below this, which is why we show a range: the base reward is what attested machines accrue for staying online regardless of traffic, and usage earnings grow with demand.
- Base rewards are paid to attested machines online ≥90% of each 5-minute settlement period, up to a fixed monthly budget — not a guarantee.
- Actual usage depends on network demand, model popularity, your reputation, and how many other providers serve the same model.
Read the technical paper
Architecture, threat model, security analysis, and economic model for private inference on distributed Apple Silicon.
Download PDF ↗