moolvylabs/Morphy-Math-1.5B

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Sep 14, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

Morphy-Math-1.5B by Moolvylabs is a 1.5 billion parameter language model with a 32768 token context length, specifically fine-tuned for advanced academic tasks. It excels in higher mathematics, calculus, linear algebra, differential equations, quantum computing, and physics. This model is optimized for complex symbolic reasoning and step-by-step problem-solving in scientific domains, designed to run efficiently on consumer-grade hardware.

Loading preview...

Morphy-Math-1.5B: Specialized for Advanced Scientific Reasoning

Morphy-Math-1.5B, developed by Moolvylabs, is the third iteration in the Morphy series, uniquely fine-tuned for exceptional performance in higher sciences. Unlike its predecessors, this 1.5 billion parameter model with a 32768 token context length is specifically optimized for complex academic datasets, making it suitable for advanced mathematical and scientific problem-solving while remaining lightweight enough for local deployment.

Key Capabilities

  • Higher Mathematics & Calculus: Solves university-level limits, improper integrals, derivatives, and infinite series with strict Chain of Thought (CoT).
  • Linear Algebra & Differential Equations: Handles non-linear differential equations, spectral theorems, matrix eigenvalues, and eigenvectors.
  • Quantum Computing & Physics: Understands advanced physics concepts, Dirac bra-ket notations, Pauli operators, and complex quantum state transformations.

Good For

  • Academic Research: Ideal for researchers and students working on advanced mathematical and physics problems.
  • Symbolic Reasoning: Excels in tasks requiring precise, step-by-step symbolic manipulation and logical deduction.
  • Local Deployment: Designed to run efficiently on consumer-grade hardware, offering powerful scientific capabilities without extensive computational resources.

Important Considerations

  • Quantization Impact: While FP16 and Q8_0 versions retain full reasoning logic, aggressive bit-compression (Q4_K_M / Q5_K_M) may show minor degradation on highly abstract, multi-variable equations.
  • Language Nuance: For maximum accuracy in complex symbolic formulas and scientific notations, prompting in English is highly recommended.

Finetunes
2 models