Native schema and guide support 128 MP. Provider workflow example and Replicate README still describe an older 8 MP contract, factor mode and default enhance_realism=true.Source for Documentation discrepancy (opens in a new tab)
Current native schema wins for native controls; Replicate has its own current schema.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 15 Oct 2026
Category
Usage
Rate
Unit
Image upscaling
Target 1–4 MP1
$0.005
image
Image upscaling
Target 5–8 MP1
$0.01
image
Image upscaling
Target 9–16 MP1
$0.02
image
Image upscaling
Target 17–32 MP1
$0.04
image
Image upscaling
Target 33–64 MP1
$0.06
image
Image upscaling
Target 65–128 MP1
$0.12
image
Rate conditions
Native integer target megapixel tier; one resulting image.
Pruna API terms retain customer ownership of input/output and permit processing to provide the service. Operational telemetry is used for security and service improvement. A universal retention period and zero-training guarantee are not established.
Current October 12 revision supports 128 MP and defaults enhance_realism=false. README still says 8 MP and true. Billing tier labels overlap at 4, 8, 16, 32 and 64 MP; exact boundary rounding is not established.Source for Documentation discrepancy (opens in a new tab)
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 15 Oct 2026
Category
Usage
Rate
Unit
Image upscaling
1-4 MP1
$0.005
image
Image upscaling
4-8 MP1
$0.01
image
Image upscaling
8-16 MP1
$0.02
image
Image upscaling
16-32 MP1
$0.04
image
Image upscaling
32-64 MP1
$0.06
image
Image upscaling
64-128 MP1
$0.12
image
Rate conditions
Replicate published target-resolution tier. Adjacent display labels overlap; exact boundary and factor-mode rounding are not established.
Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.
Caching
Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.
01 / INITIAL INPUTProcess & retain
The system processes eligible input and retains reusable state.
02 / LATER REQUESTCheck available state
Reuse depends on matching input, session state and the cache policy.
03 / CACHE HITReuse & generate
Reusing state can reduce repeated work. Any charges follow the billing rules below.
Pruna API
p-image-upscale
Not verified
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
How it works
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Minimum reusable input
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Lifetime & refresh
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
How to enable it
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Usage fields
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Billing rules
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Limits & data boundaries
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
Replicate
prunaai/p-image-upscale
Not verified
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
How it works
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Minimum reusable input
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Lifetime & refresh
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
How to enable it
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Usage fields
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Billing rules
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
Limits & data boundaries
Not collected
No verified model prompt-cache contract or caching discount. Uploaded assets and asynchronous jobs are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
What the measurements say
No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.
Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.
The record behind the model
Why it is included
Named hosted image-processing component. The exact underlying checkpoint and private-weight provenance are not established; this is explicitly a managed service record, not an additional claimed private foundation checkpoint.
