Commercial preview is described as a language diffusion world model. Research code compares SDAR and autoregressive variants; exact production checkpoint, size and lineage are not established.Source for Architecture (opens in a new tab)
Chat route rejects stream=true with HTTP 400. Native endpoint returns a complete prediction.
Generation controls
max_tokens defaults to 512 with no schema maximum; temperature defaults to 0.2, top_p to 1, top_k to 0. Nested params override top-level chat controls. Context sections are optional; interaction state and action must be nonempty strings.Source for Generation controls (opens in a new tab)
Native response contains generated content, optional reasoning, finish_reason and token usage. Chat route returns a ChatCompletion envelope. A response envelope does not guarantee schema-constrained generated content.Source for Output formats (opens in a new tab)
Access & pricing
Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Published terms retain provider platform rights and allow backup copies after account termination. Privacy policy covers submitted test cases and model outputs. A DWM-specific retention period, no-training commitment and serving region are not established; older evaluation-platform wording may not describe a negotiated DWM agreement.
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Published terms retain provider platform rights and allow backup copies after account termination. Privacy policy covers submitted test cases and model outputs. A DWM-specific retention period, no-training commitment and serving region are not established; older evaluation-platform wording may not describe a negotiated DWM agreement.
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.
Patronus DWM API
Not verified
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
How it works
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Minimum reusable input
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Lifetime & refresh
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
How to enable it
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Usage fields
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Billing rules
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Limits & data boundaries
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 Oct 2026
Patronus DWM OpenAI-compatible API
world-model
Not verified
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
How it works
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Minimum reusable input
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Lifetime & refresh
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
How to enable it
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Usage fields
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Billing rules
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
Limits & data boundaries
Not collected
No prompt-cache mechanism or billing contract established. Browser authentication storage and resubmitted transcripts are not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 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
Provider-operated prediction service under retained platform rights. The commercial alias is represented as a managed component because the served checkpoint and its weight-release status are not established. Published training code and related public checkpoints are not claimed as proprietary weights.
