VERSES® Genius™ Outperforms DeepSeek R1 Model in Code-Breaking “Mastermind” Challenge
VERSES AI Inc. (OTCQB:VRSSF) announced that its Genius™ product outperformed China's DeepSeek R1 model in the code-breaking game Mastermind. The test compared both models across 100 games, with Genius achieving superior results in multiple metrics:
- Accuracy: Genius achieved 100% success rate vs R1's 45%
- Speed: Genius solved games in 1.1-4.5 seconds vs R1's 934 seconds average
- Efficiency: Total compute time of 5 minutes vs R1's 26 hours
- Cost: $0.05 USD total cost vs R1's $38.94 USD
Overall, Genius performed 245 times faster and 779 times more cost-efficiently than DeepSeek's R1. The company emphasizes this demonstration showcases Genius' capabilities in multi-step reasoning and its potential to enhance LLM capabilities for real-world applications.
VERSES AI Inc. (OTCQB:VRSSF) ha annunciato che il suo prodotto Genius™ ha superato il modello DeepSeek R1 della Cina nel gioco di decodifica Mastermind. Il test ha confrontato entrambi i modelli in 100 partite, con Genius che ha raggiunto risultati superiori in più metriche:
- Accuratezza: Genius ha ottenuto un tasso di successo del 100% rispetto al 45% di R1
- Velocità: Genius ha risolto i giochi in 1,1-4,5 secondi contro una media di 934 secondi di R1
- Efficienza: Tempo di calcolo totale di 5 minuti contro le 26 ore di R1
- Costi: Costo totale di $0,05 USD rispetto a $38,94 USD di R1
In generale, Genius ha performato 245 volte più velocemente e 779 volte più economicamente rispetto a DeepSeek R1. L'azienda sottolinea che questa dimostrazione evidenzia le capacità di Genius nel ragionamento a più fasi e il suo potenziale per migliorare le capacità LLM per applicazioni nel mondo reale.
VERSES AI Inc. (OTCQB:VRSSF) anunció que su producto Genius™ superó al modelo DeepSeek R1 de China en el juego de resolución de códigos Mastermind. La prueba comparó ambos modelos en 100 juegos, con Genius logrando resultados superiores en varias métricas:
- Precisión: Genius alcanzó una tasa de éxito del 100% frente al 45% de R1
- Velocidad: Genius resolvió los juegos en 1.1-4.5 segundos frente a un promedio de 934 segundos de R1
- Eficiencia: Tiempo total de cómputo de 5 minutos frente a las 26 horas de R1
- Costo: Costo total de $0.05 USD frente a $38.94 USD de R1
En general, Genius operó 245 veces más rápido y 779 veces más eficiente en costos que DeepSeek R1. La empresa enfatiza que esta demostración muestra las capacidades de Genius en el razonamiento de múltiples pasos y su potencial para mejorar las capacidades de LLM para aplicaciones del mundo real.
VERSES AI Inc. (OTCQB:VRSSF)는 자사의 Genius™ 제품이 중국의 DeepSeek R1 모델을 코드 해독 게임인 Mastermind에서 능가했다고 발표했습니다. 이 테스트는 100게임에 걸쳐 두 모델을 비교했으며, Genius는 여러 지표에서 우수한 결과를 달성했습니다:
- 정확도: Genius는 100% 성공률을 달성했으며 R1은 45%
- 속도: Genius는 1.1-4.5초 만에 게임을 해결했으며 R1의 평균은 934초
- 효율성: 총 계산 시간이 5분인 반면 R1은 26시간
- 비용: 총 비용 $0.05 USD이며 R1은 $38.94 USD
전반적으로 Genius는 DeepSeek R1보다 245배 빠르고 779배 더 비용 효율적으로 작동했습니다. 이 회사는 이번 시연이 Genius의 다단계 추론 능력을 보여주며, 현실 세계 애플리케이션을 위한 LLM 능력을 향상시킬 잠재력이 있음을 강조했습니다.
VERSES AI Inc. (OTCQB:VRSSF) a annoncé que son produit Genius™ a surpassé le modèle DeepSeek R1 de Chine dans le jeu de décryptage Mastermind. Le test a comparé les deux modèles sur 100 parties, Genius obtenant des résultats supérieurs selon plusieurs critères :
- Précision: Genius a atteint un taux de réussite de 100% contre 45% pour R1
- Vitesse: Genius a résolu les jeux en 1,1-4,5 secondes contre une moyenne de 934 secondes pour R1
- Efficacité: Temps de calcul total de 5 minutes contre 26 heures pour R1
- Coût: Coût total de 0,05 USD contre 38,94 USD pour R1
Dans l'ensemble, Genius a fonctionné 245 fois plus vite et 779 fois plus économiquement que DeepSeek R1. L'entreprise souligne que cette démonstration met en avant les capacités de Genius en raisonnement par étapes et son potentiel à améliorer les capacités des LLM pour des applications dans le monde réel.
VERSES AI Inc. (OTCQB:VRSSF) gab bekannt, dass sein Produkt Genius™ das DeepSeek R1-Modell aus China im Codeknackspiel Mastermind übertroffen hat. Der Test verglich beide Modelle in 100 Spielen, wobei Genius in mehreren Kennzahlen überlegene Ergebnisse erzielte:
- Genauigkeit: Genius erreichte eine Erfolgsquote von 100% im Vergleich zu R1s 45%
- Geschwindigkeit: Genius löste die Spiele in 1,1-4,5 Sekunden, während R1 einen Durchschnitt von 934 Sekunden benötigte
- Effizienz: Die Gesamtberechnungszeit betrug 5 Minuten gegenüber 26 Stunden von R1
- Kosten: Gesamtpreis von 0,05 USD im Vergleich zu R1s 38,94 USD
Insgesamt war Genius 245 Mal schneller und 779 Mal kosteneffizienter als DeepSeek R1. Das Unternehmen betont, dass diese Demonstration die Fähigkeiten von Genius im mehrstufigen Denken zeigt und sein Potenzial zur Verbesserung der LLM-Fähigkeiten für reale Anwendungen verdeutlicht.
- Achieved 100% success rate in code-breaking challenge vs competitor's 45%
- Demonstrated 245x faster performance than DeepSeek R1
- Operated at 779x lower cost ($0.05 vs $38.94 for 100 games)
- Required only basic hardware (Mac M1 Pro) vs GPU cloud infrastructure
- None.
Demonstration of Multi-Step Reasoning by VERSES’ Genius Agent Beats China’s Top AI Model in Accuracy, Speed, and Cost Efficiency
VANCOUVER, British Columbia, Feb. 04, 2025 (GLOBE NEWSWIRE) -- VERSES AI Inc. (CBOE:VERS) (OTCQB:VRSSF) ("VERSES'' or the "Company”), a cognitive computing company specializing in next-generation intelligent agent systems, provides an update to its previously announced “Mastermind” challenge where the Company’s flagship product, Genius outperformed OpenAI’s o1-preview model. In this latest demonstration, VERSES demonstrates Genius, winning the code-breaking game Mastermind in a side-by-side comparison with China’s leading AI model, DeepSeek’s R1, which has been positioned as a disruptive, more cost-efficient large language model (LLM). Genius significantly outperformed DeepSeek’s R1 model while performing two hundred and forty-five (245) times faster and more than seven hundred and seventy-nine (779) times cheaper.
In the challenge, VERSES compared the DeepSeek-R1 model to Genius. Each model attempted to crack the Mastermind code on 100 games within up to ten guesses. Each model was given a hint for each guess and must reason about the missing part of the correct answer, requiring all six code colors to be correct to crack the code. For perspective, you can play the game at mastermindgame.org.
The highlighted results from Verses’ demonstration are below:
| Metric | Genius | R1 |
| Success Rate | ||
| Total Compute Time (100 Games) | 5m 18s (Avg 3.1s per game) | 26 hours (Avg 334s per game) |
| Total Cost for 100 Games | ||
| Hardware Requirements | Mac M1 Pro | GPU Cloud |
| $ are USD, (s) are seconds | ||
Performance Highlights:
- Accuracy and Reliability. Genius solved the code
100% of the time in a consistent number of steps. DeepSeek solved the game only45% of the time. - Speed. Genius consistently solved games in 1.1–4.5 seconds, while DeepSeek’s solve times averaged 934 seconds (approximately 15.5 mins)
- Efficiency. Genius’ total compute time for 100 games was just over 5 minutes, compared to DeepSeek’s 26 hours.
- Cost. Genius’ compute cost was estimated at
$0.05 USD for all 100 games, compared to DeepSeek’s R1 model at$38.94 USD.
In summary, Genius consistently outperformed DeepSeek’s model two hundred and forty-five (245) times faster and more than seven hundred and seventy-nine (779) times cheaper. A more detailed description and methodology of the test is on the Company’s blog at verses.ai.
“This is a good showcase of Genius’ domain-specific model’s advanced performance in a multi-step reasoning problem,” said Hari Thiruvengada, VERSES Chief Technology Officer. “Mastermind was the perfect choice for this test because it requires multi-step logical reasoning, predictive cause-and-effect understanding, and dynamic adaptation to crack the code. This exercise highlights how Genius excels in step-by-step reasoning by leveraging domain agents using a Bayesian approach and Active Inference.”
“Many AI models struggle to efficiently deliver results and as demonstrated in this challenge, Genius’ ability to perform multi-step reasoning and dynamically adjust to feedback is crucial for enabling agents that are not only more efficient but, more importantly, accurate and reliable enough to operate in dynamic real-world scenarios,” said Gabriel René, Founder and CEO of VERSES. “While this test highlights the competitive advantages of our technology, we believe Genius is also highly complementary to LLMs—enhancing their capabilities by providing the additional ‘brainpower’ needed to make AI agents smarter and more trustworthy within their domains. If DeepSeek’s R1 signals the commoditization phase of LLMs, demonstrating that general-purpose, open-source models can now be made much cheaper, we believe that this latest Mastermind test is the signal that the domain-specific models—uniquely enabled by Genius—are the missing links needed to make AI agents truly reliable. We believe this ”last mile” challenge of AI—accuracy— is the key to unlocking adoption across the Fortune 500 market and beyond,” concluded Mr. René.
Mastermind™ is a registered trademark of Pressman Inc.
About VERSES
VERSES is a cognitive computing company building next-generation intelligent software systems modeled after the wisdom and genius of Nature. Designed around first principles found in science, physics and biology, our flagship product, Genius, is a suite of tools for machine learning practitioners to model complex dynamic systems and generate autonomous intelligent agents that continuously reason, plan, and learn. Imagine a Smarter World that elevates human potential through technology inspired by Nature. Learn more at verses.ai, LinkedIn, and X.
On behalf of the Company
Gabriel René, Founder & CEO, VERSES AI Inc.
Press Inquiries: press@verses.ai
Investor Relations Inquiries
U.S., Matthew Selinger, Partner, Integrous Communications, mselinger@integcom.us 415-572-8152
Canada, Leo Karabelas, President, Focus Communications, info@fcir.ca 416-543-3120
Cautionary Note Regarding Forward-Looking Statements
When used in this press release, the words "estimate", "project", "belief", "anticipate", "intend", "expect", "plan", "predict", "may" or "should" and the negative of these words or such variations thereon or comparable terminology are intended to identify forward-looking statements and information. Although VERSES believes, in light of the experience of their respective officers and directors, current conditions and expected future developments and other factors that have been considered appropriate, that the expectations reflected in the forward-looking statements and information in this press release are reasonable, undue reliance should not be placed on them because the parties can give no assurance that such statements will prove to be correct. The forward-looking statements and information in this press release include, among other things, statements regarding the potential impacts of Deepseek R1 on the AI technology market, the expected future development of Genius and AI technology and potential adoption of AI technology.
There are risks and uncertainties that may cause actual results to differ materially from those contemplated in those forward-looking statements and information. In making the forward-looking statements in this news release, the Company has applied various material assumptions. By their nature, forward-looking statements involve known and unknown risks, uncertainties and other factors which may cause our actual results, performance or achievements, or other future events, to be materially different from any future results, performance or achievements expressed or implied by such forward-looking statements. There are a number of important factors that could cause VERSES' actual results to differ materially from those indicated or implied by forward-looking statements and information. Such factors may include, among other things, that Deepseek R1 will continue to impact the AI technology market; that Genius and AI technology generally will continue to development as currently anticipated by management; and that developments in AI accuracy will increase adoption. The Company undertakes no obligation to comment on analyses, expectations or statements made by third parties in respect of its securities or its financial or operating results (as applicable).
Additionally, forward-looking statements involve a variety of known and unknown risks, uncertainties and other factors which may cause the actual plans, intentions, activities, results, performance or achievements of the Company to be materially different from any future plans, intentions, activities, results, performance or achievements expressed or implied by such forward-looking statements. Such risks include, without limitation: that Genius may not continue to development as currently anticipated by management; that the market for AI technology may not develop as anticipated; and that developments in AI accuracy may not increase adoption. VERSES cautions that the foregoing list of material factors is not exhaustive. When relying on VERSES’ forward-looking statements and information to make decisions, investors and others should carefully consider the foregoing factors and other uncertainties and potential events. VERSES has assumed that the material factors referred to in the previous paragraph will not cause such forward-looking statements and information to differ materially from actual results or events. However, the list of these factors is not exhaustive and is subject to change and there can be no assurance that such assumptions will reflect the actual outcome of such items or factors. The forward-looking information contained in this press release represents the expectations of VERSES as of the date of this press release and, accordingly, are subject to change after such date. VERSES does not undertake to update this information at any particular time except as required in accordance with applicable laws.
FAQ
How did VERSES AI (VRSSF) Genius perform against DeepSeek R1 in the Mastermind challenge?
What were the specific performance metrics of VRSSF's Genius in the February 2025 Mastermind test?
How much more cost-efficient is VRSSF's Genius compared to DeepSeek R1?
What hardware requirements does VRSSF's Genius need compared to DeepSeek R1?
How much faster did VRSSF's Genius complete the Mastermind challenge compared to DeepSeek R1?
