Model details
Claude Opus 4.5 represents Anthropic's flagship model designed around coding, agents, and computer use as its primary strengths. The announcement positions it as the best in the world for software engineering tasks, with state-of-the-art performance on SWE-bench Verified benchmarks. Beyond technical work, the model shows meaningful improvements in deep research, document creation, and productivity tasks like working with slides and spreadsheets. The architecture is built to handle longer-running agentic tasks that require sustained attention across complex, multi-step workflows.
The model release came alongside expanded tooling including the Claude Agent SDK, which enables developers to build reliable agents that can operate across longer time horizons. Practitioners noted that the combination of Opus 4.5 and the Agent SDK functions like a waymo for work—essentially autonomous completion of complex objectives rather than just responding to individual prompts. The model became available across multiple cloud platforms at launch, including Microsoft Azure Foundry and Amazon Bedrock. The pricing structure reflects accessible rates for Opus-level capabilities, making advanced agentic AI practical for teams and enterprises looking to deploy reliable, multi-step automation across coding, research, and professional workflows.
Transparent token rates
Rates are shown per one million tokens. Combined means one million input plus one million output tokens.
Amazon Bedrock
Amazon Web Services announces that Anthropic’s Claude Opus 4.6, Claude Sonnet 4.6, Claude Opus 4.5, Claude Sonnet 4.5, and Claude Haiku 4.5 are now available through Amazon Bedrock with global cross-Region inference for customers in the Middle East, including UAE (me-central-1) and Bahrain (me-south-1). The feature rou
Amazon Bedrock
Anthropic's latest AI model is better at coding, sustaining tasks for longer and creating high-quality professional work.
Amazon Bedrock
The BenchLM aggregator page for Claude Opus 4.5 (updated September 2026) reports a 200K token context window, $5 input / $25 output pricing per million tokens, a blended price of $15, and a speed of 44 tokens per second with a 1.27-second first-token latency. The model carries a capability score of 58.4/100 against a f Category-level performance for Opus 4.5 on BenchLM includes an agentic rank of 102 of 154 (34th percentile, 43.2 score, 15 benchmarks, mixed sources), coding rank of 36 of 154 (77th percentile, 56.7 score, 5 benchmarks), knowledge rank of 60 of 184 (68th percentile, 54.0 score, 6 benchmarks), and a reasoning score of 7
Amazon Bedrock
The n8n AI Benchmark page for Claude Opus 4.5 confirms the model's spec profile: a 200,000-token context window, $0.000005 per 1K input tokens and $0.000025 per 1K output tokens, with support for file, image, and text inputs producing text output. The page describes Opus 4.5 as Anthropic's frontier reasoning model opti In n8n's benchmark evaluation, Claude Opus 4.5 achieves an overall score of 66 (19th placement), with standout marks in Structured Output (100, 1st), Hallucination (97, 2nd), Classification (19, 4th), and Tool Use (58, 3rd), while scoring 80 in Logic (7th) and 86 in Speed (14th). The benchmark page positions Opus 4.5 a
Amazon Bedrock
Discover more about what's new at AWS with Amazon Bedrock Reserved Tier available now for Claude Opus 4.5 and Haiku 4.5
