Funding
Anthropic's Claude Opus 5 Offers Longer, Complex Responses Compared to Fable 5
Anthropic has launched Claude Opus 5, which has quickly drawn attention for its longer and more detailed responses compared to the previous version, Opus 4.8. Users have noted that the new model not only offers a 1 million token context window but also has a maximum output token capacity of 128,000 tokens.
In terms of pricing, Opus 5 remains consistent with its predecessor, charging $5 per million input tokens and $25 per million output tokens. This pricing model is significantly lower than that of Fable 5, which costs about 45 times more for API access.
Opus 5 is now the default model on Claude Max and the strongest option available on Claude Pro, featuring a Fast mode for quicker responses. However, its tendency toward verbosity requires users to prompt for brevity, marking a shift in user interaction compared to Opus 4.8.
In contrast, Fable 5 is characterized by concise responses that often include validation. While it scores slightly better on broader intelligence metrics, the cost-effectiveness of Opus 5 raises questions about its competitive positioning in the AI landscape.
New Developments in AI Competition
DeepSeek, a leading Chinese AI lab, has formed a new "Harness Team" to develop AI agents aimed at competing with Anthropic’s Claude Code. This initiative highlights DeepSeek's ambition to strengthen its position in the AI agent market.
The team began assembling in May 2026 and is actively hiring for key roles in Beijing, including a Product Manager and R&D Engineer, to support the "DeepSeek Code Harness" project.
Although the product is not yet in the shipping phase, this move indicates a strategic push to enhance software orchestration capabilities and compete with established players like OpenAI Codex and GitHub Copilot.
Market dynamics suggest that DeepSeek's initiative may exert moderate pressure on Anthropic's perceived lead in the market.
Observers should keep an eye on DeepSeek’s hiring progress and any announcements regarding the "DeepSeek Code Harness" as these developments could signal increased competition for Anthropic.
FAQ
What are the key features of Claude Opus 5?
Claude Opus 5 offers a 1 million token context window, a maximum output token capacity of 128,000 tokens, and is designed to provide longer and more detailed responses compared to its predecessor, Opus 4.8.
How does the pricing of Claude Opus 5 compare to Fable 5?
Opus 5 charges $5 per million input tokens and $25 per million output tokens, which is significantly lower than Fable 5, which costs about 45 times more for API access.
What is the default model for Claude Max?
Claude Opus 5 is now the default model on Claude Max and is the strongest option available on Claude Pro.
Does Claude Opus 5 have a mode for quicker responses?
Yes, Claude Opus 5 features a Fast mode that allows for quicker responses, although it may require users to prompt for brevity due to its tendency toward verbosity.
How does Fable 5 differ from Claude Opus 5 in terms of response style?
Fable 5 is characterized by concise responses that often include validation, while Claude Opus 5 provides longer and more detailed answers, marking a shift in user interaction.