Apollo (AI Physics Models)
LeaderApollo is an open model family of AI physics models that provides pretrained checkpoints and reference workflows for surrogate modeling and near-real-time simulation across multiple physics domains.
Apollo is an open model family of AI physics models that provides pretrained checkpoints and reference workflows for surrogate modeling and near-real-time simulation across multiple physics domains.
Published May 19, 2026
| Starting Price | Not publicly listed | Not publicly listed (quote-based; trial available) | Not publicly listed (enterprise licensing) |
| Typical Use Case | Teams embedding open physics AI models into simulation stacks; GPU-first organizations | CAE organizations seeking packaged AI surrogate workflows (GUI + integration with HyperWorks) | Enterprises standardized on Simcenter tools seeking AI/ML acceleration |
| Key Strengths | Alignment with NVIDIA ecosystem (GPUs, frameworks); multi-domain model family and reference workflows | Productized user experience; CAE-focused documentation and workflow tooling | Deep integration within Siemens engineering toolchain and CAE credibility |
| Key Weaknesses | Limited independent reviews; unclear packaging/pricing/licensing; integration and validation work required by buyer | Potential vendor dependency on Altair tooling; requires historical data quality | Can be complex and costly; AI features may be distributed across modules |
—