Hajra Naz
Anthropic has introduced Claude Opus 5, calling it its most capable AI model yet. The company says the new model delivers stronger performance while keeping the same pricing as its previous generation.
The launch comes as businesses pay closer attention to AI costs. Many companies now want better value instead of simply choosing the most powerful model available.
Anthropic says Claude Opus 5 offers frontier-level AI performance at half the price of Claude Fable 5. Despite the performance jump, pricing remains unchanged from Claude Opus 4.8. The model costs $5 per million input tokens and $25 per million output tokens.
According to Anthropic, Opus 5 performs better than Fable 5 on coding tasks and knowledge-based evaluations. The company says it is designed for everyday business use rather than occasional high-end workloads.
One of the biggest upgrades is the model’s reasoning.
Earlier versions required developers to enable extended reasoning manually. Claude Opus 5 makes that decision on its own. It automatically chooses when extra thinking is needed before generating a response.
Anthropic says this should improve answer quality without requiring additional setup. However, developers may need to adjust their output-token limits because reasoning and responses now share the same token budget.
The company also introduced a new reasoning scale. Users can select different effort levels, including a new Max mode for the most complex tasks. Anthropic notes that disabling reasoning is only supported at lower effort levels. Requests to disable it in higher reasoning modes will return an error.
Read More: Anthropic Upgrades Claude Voice Mode With Smarter AI Models
Claude Opus 5 also brings major improvements for developers.
Anthropic says the model can complete large coding projects across multiple files with fewer unfinished sections. It also performs stronger code reviews and detects software bugs with fewer false positives.
The company says vision capabilities have improved as well. Claude Opus 5 can better understand charts, documents, spreadsheets, user interfaces, and presentation slides.
Another major feature is its 1-million-token context window. This allows the model to process much larger amounts of information during a single conversation.
Several smaller changes are also included. Claude Opus 5 writes longer responses by default. It explains its progress more clearly during agentic tasks. The model also verifies its own work automatically, reducing the need for extra verification prompts.
Anthropic says developers should remove older prompts that ask Claude to double-check its work, as the model now performs those checks on its own.
The company also stressed that Claude Opus 5 is not its most advanced model for sensitive areas such as offensive cybersecurity or biology research. Those capabilities remain behind its specialized Mythos 5 model.
The release comes as competition in enterprise AI continues to grow. Anthropic is competing with OpenAI, Microsoft, Google, Amazon, and several Chinese AI companies, all of which are introducing lower-cost models.
According to Anthropic’s Head of Product Management for Research, Dianne Penn, businesses are looking for real value instead of simply lower prices. She said a cheaper model is only useful if it can deliver the same level of quality.
Claude Opus 5 is available now through the Claude API, AWS, Google Cloud, and Microsoft Foundry.
