Foam-Agent
LLM based multi-agent framework that automates OpenFOAM CFD workflows from natural language prompts, handling meshing, case setup, execution and error correction, and exposes the workflow as an MCP server.
What it does and how it connects
Foam-Agent automates OpenFOAM simulations from a natural language prompt or an existing case. Its agents plan the case, prepare meshes and input files, run locally or through Slurm, repair errors, and optionally visualize results.
It targets OpenFOAM Foundation v10 by default, with a separate native ESI/OpenCFD v2006 target, and retrieves tutorial references through FAISS indices for the selected target. Model backends include OpenAI, Codex subscription, Anthropic, Bedrock, Ollama, and DeepSeek, and the workflow is exposed through MCP integration for Claude Code, Cursor, and other MCP clients.
The recommended install is Docker, which bundles Foundation OpenFOAM v10, Python dependencies, tutorial indices, and the default embedding weights, so no host Conda, OpenFOAM, or GPU is needed. The project is published as a paper (arXiv 2505.04997) in Computer Methods in Applied Mechanics and Engineering, per the README and repo description.
What engineers report
Great tool! But: Why are this old versions here supported? OF V14 and ESI 26.06 are actual.
2026-11-01Read the original
Alternatives
Compare side by side- Ansys SimAI
A commercial physics-agnostic AI platform for simulation that trains models on existing simulation data to predict new design performance in minutes.
- Monolith AI
A commercial no-code machine learning platform for engineering that learns from test and simulation data to predict design performance.
- PhysicsX
A commercial AI-native engineering platform that uses physics AI models and Large Physics Models to evaluate simulations in seconds.
Sources
Not verified
- Pricing: none stated on the repo.
- Related Error Index entries: no matching entry verified, since no OpenFOAM entries exist in the live map.
- A named individual builder beyond the repository owner organization csml-rpi
Last verified November 6, 2026. Listings are informational. The evidence score measures public, verifiable evidence and is not an endorsement or a compatibility claim.