Anthropic Claude Opus 4.8, the flagship of the Claude 4 family. Building on Opus 4.7 with sharper judgment, more reliable agentic behavior, and notably improved honesty: it is far less likely to make unsupported claims or to let flaws in its own code pass unremarked. Top-tier reasoning and writing for the most critical analyses and the content you ship to customers. Same pricing as Opus 4.7. If you need to search the web for information, please turn on the "Web Search" switch. Prompt example 1: (Upload a residential lease PDF) Review it clause by clause, flag terms unfavorable to the tenant, explain the risks, and suggest revisions. Prompt example 2: Outline the literature review for an academic paper on how remote work affects team collaboration, with section titles, key arguments, and the types of studies to cite. Prompt example 3: Write a short story of about 2,000 words about an old watchmaker repairing his last watch before retiring, with delicate prose and a surprising twist at the end. Prompt example 4: Our e-learning platform is expanding to Europe and the US. Compare keeping data in Taiwan with regional deployment on latency, consistency, privacy laws, and upkeep, and note what needs testing. Prompt example 5: Write a 1,500-word academic essay presenting both sides of whether AI should be listed as a co-author on research papers, ending with your own position and reasons. Prompt example 6: Judge each part of this claim as supported, unsupported, or wrong, and say clearly what you are unsure of: Humans use only 10 percent of their brains, so brain training can unlock the rest. Prompt example 7: Write a 1,200-word graduation speech for an elementary school principal about to retire, reflecting on 30 years in education, sincere but not sentimental, ending with three pieces of advice. Prompt example 8: (Upload a survey-based research paper PDF) Evaluate its sampling, questionnaire reliability and validity, and statistical methods, point out likely biases, and suggest improvements. Prompt example 9: (Upload a 300-page database systems textbook PDF) Turn it into chapter-by-chapter study notes, each with core concepts, common misconceptions, and two self-test questions. Prompt example 10: Write a 3,000-word feature for a company's 50th anniversary booklet tracing a hardware store's growth from a corner shop to a chain brand, in elegant prose across four chapters. Maximum context window (input and output combined): about 1,000,000 tokens. Maximum output per reply: about 64,000 tokens (model reasoning, if any, counts toward this limit). Maximum total upload size per request: 512MB (all files and message content combined). Accepted document formats: pdf, docx, xlsx, pptx, txt, csv (file content is provided to the model as extracted text). Accepted image formats: gif, jpeg, jpg, png, webp. Animated GIF images are analyzed using only the first frame. Maximum number of images: 600. When uploading more than 20 images, each side must be at most 2,000 pixels. Maximum size per image: 10MB (base64-encoded). Maximum total image size per message: 20MB (text and history share the same request size limit; uploads close to the limit may still be rejected). Image dimensions: 200 to 8,000 pixels per side. Output content formats: Text/Image Pricing: High Continuous conversation: Supported AI can make mistakes. Please verify important information.