DeepSeek models on Enclave42
Enclave42’s model catalogue can include DeepSeek-family models, served through the same confidential-inference API as the rest of the catalogue. Availability changes — the catalog is the current source of truth, not this page.
DeepSeek on Enclave42
When listed, DeepSeek-family models are accessed the same way as any other model in the catalogue: one API, a model ID, and plan-based access. We don’t fix pricing, context window or availability on this page, because catalog data changes — check the live model catalog for what’s currently offered and at what price.
Confidential by default
Models in the Enclave42 catalogue run as confidential workloads, processed inside an Intel TDX trust domain on an NVIDIA GPU, with customer-verifiable attestation.
Benchmarks
We publish measured latency and throughput figures for select catalog models. See the benchmarks page for the current measured set.