Update README.md

README.md

@@ -65,14 +65,16 @@ extra_gated_button_content: Agree and Request Access
 
 ## Model Description
 
-This is **Ex0bit/GLM-4.7-Flash-PRISM** 
-- PRISM (Projected Refusal Isolation via Subspace Modification) — State-of-the-art over-refusal/propaganda removal from LLMs
+This is **Ex0bit/GLM-4.7-Flash-PRISM**
 
-**GLM-4.7-PRISM:** Unrestricted GLM-4.7 Model Access Access GLM-4.7-PRISM, an abliterated version of ZAI's flagship 358B parameter model with over-refusal and propoganda mechanisms removed. 
+**GLM-4.7-Flash-PRISM:** Unrestricted GLM-4.7-Flash Model Access
 
-**What You Get:**  
-- 358B Parameter MoE Architecture — Full-scale best-in-class agentic foundation model with ~32B active parameters.
-- PRISM (Projected Refusal Isolation via Subspace Modification) State-of-the-art abliteration technique that removes over-refusal and propaganda behaviors from the model while preserving/enhancing capabilities
-- 131K Context Window — Extended context for complex tasks
-- Preserved Thinking — Multi-turn reasoning that persists across conversations
-- Strongest in-class Benchmarks — 73.8% SWE-bench, 87.4% τ²-Bench, 95.7% AIME 2025
+Access GLM-4.7-Flash-PRISM, an abliterated version of ZAI's efficient 30B-A3B MoE model with over-refusal mechanisms removed.
+
+**What You Get:**
+
+- **30B-A3B MoE Architecture** — Lightweight yet powerful Mixture-of-Experts model with 30 billion total parameters and ~3 billion active per token for fast, efficient inference
+- **PRISM (Projected Refusal Isolation via Subspace Modification)** — State-of-the-art abliteration technique that removes over-refusal behaviors while preserving capabilities
+- **128K Context Window** — Extended context for complex tasks and large codebases
+- **Interleaved & Preserved Thinking** — Multi-turn reasoning that persists across conversations with per-turn thinking control
+- **Strong In-Class Benchmarks** — 91.6% AIME 2025, 79.5% τ²-Bench, 59.2% SWE-bench Verified, 75.2% GPQA