SuperMoney’s Sense AI Achieves 99.4% Accuracy on CFP® Financial Planning Benchmark

SuperMoney has released performance data for Sense AI, its agentic financial assistant, demonstrating a 99.4% median accuracy rate against a rigorous 1,000-question benchmark. For fintech professionals, this represents a significant milestone in the reliability of AI-driven advice, suggesting that large language models are becoming increasingly capable of handling complex, regulated financial planning domains with minimal error.

What was announced

SuperMoney conducted a comprehensive evaluation of Sense AI, the core artificial intelligence component within its agentic finance application. The benchmark consisted of 1,000 questions specifically authored to align with the CFP® Board’s 2026 Principal Knowledge Topics. This blueprint serves as the foundational curriculum for the CFP® certification exam, covering eight distinct domains: Professional Conduct and Regulation, General Principles of Financial Planning, Risk Management and Insurance Planning, Investment Planning, Tax Planning, Retirement Savings and Income Planning, Estate Planning, and the Psychology of Financial Planning.

To ensure the statistical validity of the results, SuperMoney ran the entire 1,000-question bank seven times. The testing was performed on the live production system utilized by SuperMoney members, rather than a controlled laboratory environment. Each question was answered in an isolated session to prevent the model from utilizing context from previous queries, thereby ensuring each response was generated independently. The results showed a high degree of consistency, with a standard deviation of only 0.11 percentage points.

Domain-specific performance was notably high, with Sense AI achieving a perfect 100% score across all seven runs in the Insurance, Retirement, and Psychology of Financial Planning categories. The lowest performing area was Tax Planning, which still maintained a high accuracy threshold of 97.97%. While the benchmark does not use copyrighted CFP® exam materials, it represents a targeted effort to measure AI competence against the industry's highest professional standards.

"In personal finance, confident and wrong is the most expensive combination there is. We think any company making claims about financial competence should be willing to measure it."

Miron Lulic, founder & CEO, SuperMoney.

The companies involved

SuperMoney is a financial technology firm focused on democratizing access to financial literacy and planning tools through its agentic finance platform. The company operates in a competitive market where the integration of artificial intelligence into consumer-facing applications is becoming a standard requirement rather than a luxury. By positioning its "Sense AI" assistant as a tool capable of navigating the complexities of the CFP® Board’s knowledge domains, SuperMoney is attempting to bridge the gap between automated self-service tools and professional human advisory services.

Led by CEO and Founder Miron Lulic, the company emphasizes transparency in AI performance, particularly regarding the "hallucination" risks associated with large language models in the financial sector. Unlike traditional lead-generation sites that simply aggregate financial products, SuperMoney’s recent pivot toward agentic finance suggests a broader ambition to act as a proactive partner in a user's financial life. The company remains an independent entity in the fintech space, leveraging its proprietary technology to compete with both established financial institutions and newer, AI-native startups vying for the role of the consumer's primary financial interface.

What this means

This benchmark is a direct challenge to the "black box" nature of most financial AI assistants. By hitting a 99.4% median accuracy, SuperMoney is signaling that the era of AI "hallucinations" being an acceptable risk in fintech is drawing to a close. The perfect scores in retirement and insurance planning are particularly significant, as these areas often involve the most rigid regulatory frameworks and long-term consequences for consumers.

However, the 97.97% score in Tax Planning—while high—reminds the industry that tax law remains the most difficult frontier for AI to master. Competitors will now be under pressure to release their own audited accuracy scores rather than relying on vague marketing claims. The next step for the industry will be moving from answering questions to executing complex, multi-step financial tasks with the same level of verified precision.