Discover the Best AI Tools

Provides AI agents with symbolic and numerical computation, mathematical claim checks and SVG plots through MCP or CLI. Requires Node.js and a supported expression format; computer algebra checks do not constitute general formal proofs.

mcp-gnu-units exposes offline dimensional unit conversion, search and definitions using a bundled GNU Units database and a Python engine. It is a prerelease installed from source rather than a published PyPI package, under GPL-3.0-or-later. Conversion coverage depends on the engine and included definitions.

gnomon-mcp offers deterministic batch date, timezone, mathematical, statistical and unit calculations through Python MCP tools. Working-day arithmetic treats Monday through Friday as workdays and does not apply regional holidays. Natural-language dates and units still require unambiguous inputs.

Polymath MegaBlaster MCP provides local Node.js tools for mathematical expressions, statistics, dates, units, subnets and token counting. Runtime calculations need no API key. Precision and tokenizer coverage depend on the selected operation; geographic distance uses an Earth model and is not an exact measurement.

Mathlas gives AI clients mathematical tools for numerical checks, constant and sequence identification, theorem search and Lean proof checking. It contains no internal LLM; some tools require local OEIS data or Lean, while theorem retrieval can use external services.

MathMethods MCP exposes numerical root finding, integration, differentiation, interpolation and differential-equation solvers to AI clients. It provides numerical approximations under each method's assumptions, rather than symbolic proofs or guaranteed convergence.

mcp-abacus provides expression evaluation, numerical solving and curve fitting through MCP, using selected fixed-point, floating-point, rational or complex modes. Results carry precision information. Supported operations vary by mode and solver; an exact label concerns that calculation, not the correctness of the submitted formula or assumptions.

Language-model assistant for theorem proving in Lean.

Geometry theorem prover combining a language model with a symbolic deduction engine.

An educational MCP server for math operations, statistics, visualization, and persistent workspaces.