Step 1
Choose your Mac
We use your chip and memory to estimate how fast each model can run. Your choice is saved in this browser.
How accurate is this speed estimate?
It is a broad estimate for planning. A real speed test on your Mac will be more accurate. Models that need more memory than your Mac has are left out.
Choose a guide
What do you want to find out?
You can ask for our best choices or check models that you already have in mind.
New estimates are paused while network activity changes quickly.
Rankings are paused because this page has not received new data for more than 3 minutes.
The last update remains visible. Wait for new data before you use a ranking or recommendation.
Step 2
Choose the models you want to estimate
Select one or more models. We will check whether they fit and estimate them together.
24-hour estimate for your selected models
This is the value of completed work at Darkbloom’s listed token prices, not a promised payout. Compare the result using activity from the last hour, last 24 hours, or last 7 days.
This is one combined estimate. The same Mac time is not counted more than once.
Step 2
Model options for your Mac
We rank the number of ready models selected above. Several-model rankings include only sets seen in usable public data.
Building your ranking…
We are combining the selected Mac speed with the latest model activity and prices.
This ranking will update when you change your Mac.
How this order works
For one model, the option with the highest low estimate comes first. A model can have a large high estimate and still appear lower when its low estimate is near zero.
For several models, each complete set gets a new combined estimate. The order uses the low end of that estimate. Time providers with this set were working uses one shared time total, so each provider hour is counted only once.
See all observed models
Use this table when you want to compare every model seen in the public data. The options above are enough for most people.
| Model | Time providers were working | Likely working-time range | Time requests were waiting | Darkbloom’s listed price | Estimated speed on your Mac | Estimated work value range | Estimate confidence | Result status |
|---|---|---|---|---|---|---|---|---|
| Loading the latest data… | ||||||||
Estimated work value means the value of completed work at Darkbloom’s listed token prices. It uses recent activity, expected types of work, and the estimated speed of your Mac. It is not a promised payout.
Optional detail
Review one model
Choose a model to see its recent activity and how we made its estimate.
This choice changes only the review below. It does not change your model guide above.
Selected model
Loading this model…
Please wait while we load the latest activity.
Based on Darkbloom’s listed prices. This is not a promise of payment.
See how this estimate was made
Estimate details
Select a model
Choose a model above to see what was used in its estimate.
Use my own measured speed
Optional live view
What this Mac is doing now
See Darkbloom activity and recently completed work. These values update each second.
Connecting
Waiting for the first local sample…
Completed output over time
Request-finish token bursts, up to the last 60 state updates. This is not a live decode trace.
Current activity
Connecting
Process load identifies heavy work even when no request has finished.
GPU memory
—
Models ready
Waiting…
True live decode and prefill tok/s are not exposed.
Darkbloom adds the available token counter when a coordinator request finishes. The dashboard does not convert those bursts into a false live rate.
How the estimate is calculated
We first measure how much time providers were working on each model. We then combine that activity with Darkbloom’s listed price and the estimated speed of your Mac.
A model is recommended only when its lowest estimate is higher than the highest estimate for every other model. If the ranges overlap, we say that there is no clear first choice.
How provider work time is measured
time providers were working ÷ total time providers were ready for that model
We use only updates where the two public provider counts agree. We leave out long gaps.
How the work value is estimated
The estimate combines the input price, output price, expected input/output mix, Mac speed, selected time period, and recent provider work time.
Work patterns used in the estimate
The public data does not show the exact input and output mix for each model.
Limits of this estimate
- This estimate cannot explain why one provider did or did not get work.
- Darkbloom’s listed price is not proof of the amount paid to a provider.
- Each exact model version stays separate.
- The Mac speed values are estimates until you replace them with your own test results.
Data and privacy
Network tokens in the last 24 hours
Not available Extra network context. It does not change the model result.How a recommendation is chosen
Estimated ranges must not overlap We do not recommend one option when the possible results overlap.Your saved information
This browser only Your Mac choice and measured speeds are not sent to the data collector.