Claude Opus 4.5

Claude Opus 4.5 is a frontier large language model developed by Anthropic, released in November 2025. It excels in software engineering, agentic tasks, reasoning, and multimodal understanding, with state-of-the-art performance on benchmarks like SWE-bench Verified (80.9%), Terminal-Bench (59.3%), and ARC-AGI-2 (37.6%). The model supports up to 1M tokens (native context window) and demonstrates strong capabilities in tool use, computer use, and browser automation. Claude Opus 4.5 was trained on a proprietary mix of publicly available internet data (up to May 2025), private datasets, opted-in user data, and synthetic data, with deduplication, classification, and filtering applied. It underwent RLHF (Reinforcement Learning from Human Feedback) and AI feedback fine-tuning to align with Claude’s Constitution (helpful, honest, harmless). The model is multilingual and text-only. Safety evaluations show strong alignment, with a 99.78% harmless response rate on violative requests and low over-refusal rates (0.23%) on benign prompts. It was deployed under AI Safety Level 3 (ASL-3) protections, with no evidence of crossing CBRN-4 or AI R&D-4 thresholds. Notably, it improves on honesty, agentic safety, and multi-turn conversational robustness compared to prior models.