d4rkninja/tanpo-finance

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.2BQuant:BF16Context Size:32kPublished:Sep 18, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

d4rkninja/tanpo-finance is a compact 1.2 billion parameter language model, fine-tuned from LiquidAI/LFM2.5-1.2B-Instruct with a 32,768-token context length. This specialist model is optimized for financial workflows, including unit economics, runway planning, pricing analysis, and SaaS metrics. It is designed for local, inexpensive deployment, offering focused assistance for business finance tasks.

Loading preview...

Overview

Tanpo Finance is a compact, 1.2 billion parameter language model developed by d4rkninja, fine-tuned from LiquidAI/LFM2.5-1.2B-Instruct. It features a 32,768-token context length and is part of the Tanpo family of domain-specialized business models. Unlike general-purpose LLMs, Tanpo Finance is specifically designed for financial workflows, enabling private and low-cost inference for focused business needs.

Key Capabilities

  • Financial Planning: Excels at unit economics, runway and cash planning, pricing, and margin analysis.
  • SaaS Metrics: Supports analysis of SaaS metrics, cohort revenue, and forecast scenarios.
  • Efficiency: Optimized for local, edge, and inexpensive deployment, making it suitable for environments where larger models are impractical.
  • Performance: Achieved a 98.4% overall rubric score in internal evaluations, a +5.4 percentage point improvement over its base model on finance-specific tasks.

Best For

  • Analyzing unit economics, contribution margin, and LTV/CAC.
  • Developing runway, burn, cash-bridge, and stress-scenario plans.
  • Drafting pricing, packaging, margin, and budget tradeoff analyses.
  • Reviewing SaaS metrics, cohort revenue, and finance operating forecasts.

Limitations

  • Specialized Scope: Quality degrades outside its finance and business-workflow domain.
  • Scale: As a 1.2B parameter model, it has limited world knowledge and arithmetic reliability compared to larger models.
  • Accuracy: Outputs may contain incorrect formulas or assumptions; human verification with source systems and finance professionals is crucial.
  • Specific Weaknesses: Trails the base model in cost_structure evaluation and identity_guardrails scores 90.0%.

Finetunes
3 models