Darkbloom
Ranked 4 of 12 LLM inference APIs by recommenders
Pays Mac owners for idle compute to run AI inference, praised for the economic idea and hardware-attested privacy angle. Critics doubt the privacy claims and the earnings math.
Pays Mac owners for idle compute to run AI inference, praised for the economic idea and hardware-attested privacy angle. Critics doubt the privacy claims and the earnings math.
58 people gave an opinion
- recommend it
- 19
- criticize it
- 39
- approval share
- 33%
Recommended in 1 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: not recommended
- April 2026: recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 19 people recommend itNeeds 10 or more, met.
- 0.4 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 1 of 13 monthsNeeds 4 or more, not met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Idle Macs earn money
“Part of a practical economic solution and may not be perfect so far is to use local AI for people to get paid for inference”
Comment from 19 Apr 2026 in “The Unabomber Manifesto in 2026 – Scary Accurate” “this looks like a great idea and might have a chance at solving the economic issue with running nodes for cheap inference”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “It doesn't cost you anything to have it running, and if you get requests, you make money.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” - Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
Seen as a sensible, good project
“That's a good, valid approach and should work”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “That solution actually makes great sense.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “It's a good project that makes sense.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” - Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
Hardware-attested privacy approach is interesting
“the hardware-attested privacy path is the interesting part of this, but the economic side has a quieter risk the thread has not named”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “would have at least some credibility if there was a way to remotely attest a Mac's boot sequence and TCC configuration”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
What people warn about
Each mark is one comment
Privacy claims don't hold up
“would have at least some credibility if there was a way to remotely attest a Mac's boot sequence and TCC configuration”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “there's no hardware enclave for third-party code available to do the type of attestation that would be necessary”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “I am not entirely sure they understand that System Integrity Protection and Secure Boot can be turned off.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” - Comment from 19 Apr 2026 in “Darkbloom – Private inference on idle Macs”
- Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
- Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
Earnings math doesn't add up
“Because their numbers don't work out. When you do the math on token cost versus inference speed, you get something that barely breaks even even with cheap power.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “they subtract 12Ws of idle power, because they are assuming your machine is idling 24/7”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” Hardware wear and thermal strain
“Idle chips are mostly powered down, non-idle ones are working and that causes real measurable wear and tear on the silicon.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “You're not taking into account the thermal strain on the machine, though.”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” Few paying customers, little demand
“essentially unusable as a provider unless they perform some serious lift to get actual paying customers”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs” “they do not generate inference requests to their network themselves to motivate early adopters”
Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
Other LLM inference APIs people recommend
The four most recommended other LLM inference APIs, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Cerebras Inference | 109 | 51 | 68% | 13 | — | Approved |
| Groq API | 58 | 23 | 72% | 12 | — | Approved |
| Tinfoil | 21 | 10 | 68% | 10 | — | Approved |
| DarkbloomThis page | 19 | 39 | 33% | 1 | — | — |
| Synthetic | 18 | 1 | 95% | 11 | — | Approved |
What people said
40 of the 68 comments that mention it, newest first. Each links to the original comment.
“Everything they mention can be bypassed, so determined person can start stealing data from the network.”
“Part of a practical economic solution and may not be perfect so far is to use local AI for people to get paid for inference”
“Idle chips are mostly powered down, non-idle ones are working and that causes real measurable wear and tear on the silicon.”
“essentially unusable as a provider unless they perform some serious lift to get actual paying customers”
“kept having issues getting their server to properly serve the model”
“the hardware-attested privacy path is the interesting part of this, but the economic side has a quieter risk the thread has not named”
“it doesn't inspire confidence to see that their profile is unsigned”
“It's a good project that makes sense.”
“I wish this was self hostable, even for a license fee.”
“you can't seem to buy credits right now, it says "This page couldn’t load"”
“I think it's a mistake to assume any kind of privacy for this service.”
“Actually more useful than Bitcoin.”
“Because their numbers don't work out. When you do the math on token cost versus inference speed, you get something that barely breaks even even with cheap power.”
“You're not taking into account the thermal strain on the machine, though.”
“Interesting to see an offering with this heritage proposing flat earnings rates”
“I really want this to succeed”
“their privacy posture at the cornerstone of their claims is snake oil and has gaping holes in it”
“your requests are still going through the coordinator of the system here”
“would have at least some credibility if there was a way to remotely attest a Mac's boot sequence and TCC configuration”
“Basically that computer is theirs now. So don't plan on just handing over your laptop temporarily”
“there's no hardware enclave for third-party code available to do the type of attestation that would be necessary”
“It doesn't cost you anything to have it running, and if you get requests, you make money.”
“there's no actual way to verify the contents of the running binaries”
“Also the benchmark just doesn't work.”
“I am not entirely sure they understand that System Integrity Protection and Secure Boot can be turned off.”
“they do not generate inference requests to their network themselves to motivate early adopters”
“latest (v0.3.8) tar doesn't contain image-bank or gRPCServerCLI dependencies so installer fails”
“They lost me with just one microcopy”
“I still think it's a solid economic model that benefits low income countries the most”
“Apple should build this, and start giving away free Macs subsidized by idle usage.”
“Fails to download the image model. The audio/tts model fails to load.”
“they subtract 12Ws of idle power, because they are assuming your machine is idling 24/7”
“I'd say it's not worth it. But the idea is cool.”
“This is a great use case for it.”
“Generate images requested by randoms on the internet on your hardware.”
“anyone sending their prompt/data to an external server should assume there will be some level of retention”
“Cool idea. Just some back-of-the-envelope math here (not trusting what's on their site)”
“That solution actually makes great sense.”
“That's a good, valid approach and should work”
“this looks like a great idea and might have a chance at solving the economic issue with running nodes for cheap inference”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.