Update README.md

README.md

@@ -31,6 +31,7 @@ license: mit
 
 **Highlights**
 - Half the size of SOTA models like QWQ-32b and EXAONE-32b and hence **memory efficient**.
+- It consumes a fraction of thinking tokens as compared to other SOTA models. Hence mot only is it memory efficient but **also faster** 🚀🚀🚀 
 - On par or outperforms on tasks like - MBPP, BFCL, Enterprise RAG, MT Bench, MixEval, IFEval and Multi-Challenge making it great for **Agentic / Enterprise tasks**.
 - Competitive performance on academic benchmarks like AIME-24 AIME-25, AMC-23, MATH-500 and GPQA considering model size.
 
@@ -48,6 +49,12 @@ Evaluations were conducted using [lm-eval-harness](https://github.com/EleutherAI
 
 ![image/png](https://cdn-uploads.huggingface.co/production/uploads/63d3095c2727d7888cbb54e2/psk628cePXQZ9AghKuI5u.png)
 
+
+
+
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/63d3095c2727d7888cbb54e2/jmvBwsJAIZVTP0xtWSTIz.png)
+
+
 ---