Anthropic announced on April 9 that it would正式地正式引入 the “Advisor Strategy” into the Claude Platform. Developers can now have Opus play an advisory role through the Messages API, pairing Sonnet or Haiku as the executor, achieving near-Opus-level intelligence in a single API request while significantly lowering costs. This capability is currently offered in the form of a public beta.
How it works: executors automatically consult Opus when they encounter difficult problems
The core idea behind the Advisor Strategy is to separate reasoning from execution. After developers add an advisor tool in their Messages API call, when a Sonnet or Haiku agent faces a difficult decision during task execution, it will automatically consult Opus, obtain planning guidance, and then continue executing. The entire process is completed within a single API request, with no extra orchestration required from the developer.
This architecture is similar to the division of labor between senior consultants and operators in a team—Opus handles high-level strategic judgment, while Sonnet or Haiku handles actual operations. The two collaborate within the same workflow.
Hands-on results: better performance at lower cost
According to the evaluation data Anthropic published, on the SWE-bench Multilingual benchmark, Sonnet combined with the Opus advisor performed 2.7 percentage points better than using Sonnet alone, while the cost per task actually dropped by 11.9%.
The reason costs decreased is that most tasks are still handled by the cheaper Sonnet or Haiku. Only at critical decision points is Opus called, avoiding the high expense of using Opus throughout the entire process.
What it means for developers
The Advisor Strategy provides developers building AI agents with a new cost-efficiency trade-off option. In the past, developers had to choose between “using Opus end-to-end but at high cost” or “using Sonnet end-to-end, but risking mistakes when encountering complex problems.” Now, it’s possible to get the advantages of both through a hybrid architecture.
The community’s reaction has been enthusiastic. Some developers likened it to the “call for help” lifeline from the “Millionaire” TV show—executors work independently most of the time, and only turn to the experts when they encounter truly difficult problems.
This article, Anthropic launches the Advisor Strategy: Opus as an advisor paired with Sonnet for execution, improving performance and reducing costs by 12%, first appeared on 鏈新聞 ABMedia.
Disclaimer: The information on this page may come from third parties and does not represent the views or opinions of Gate. The content displayed on this page is for reference only and does not constitute any financial, investment, or legal advice. Gate does not guarantee the accuracy or completeness of the information and shall not be liable for any losses arising from the use of this information. Virtual asset investments carry high risks and are subject to significant price volatility. You may lose all of your invested principal. Please fully understand the relevant risks and make prudent decisions based on your own financial situation and risk tolerance. For details, please refer to
Disclaimer.
Related Articles
DeepSeek Slashes Input Cache Prices to 1/10 of Launch Price; V4-Pro Drops to 0.025 Yuan per Million Tokens
Gate News message, April 26 — DeepSeek has reduced input cache prices across its entire model lineup to one-tenth of launch prices, effective immediately. The V4-Pro model is available at a limited-time 2.5x discount, with the promotion running through May 5, 2026, 11:59 PM UTC+8.
Following both re
GateNews2h ago
OpenAI Recruits Top Enterprise Software Talent as Frontier Agents Disrupt Industry
Gate News message, April 26 — OpenAI and Anthropic have been recruiting senior executives and specialized engineers from major enterprise software companies including Salesforce, Snowflake, Datadog, and Palantir. Denise Dresser, former CEO of Slack under Salesforce, joined OpenAI as chief revenue of
GateNews2h ago
Baidu Qianfan Launches Day 0 Support for DeepSeek-V4 with API Services
Gate News message, April 25 — DeepSeek-V4 preview version went live and open-sourced on April 25, with Baidu Qianfan platform under Baidu Intelligent Cloud providing Day 0 API service adaptation. The model features a million-token extended context window and is available in two versions: DeepSeek-V4
GateNews8h ago
Stanford AI course combined with industry leaders Huang Renxun and Altman, challenging to create value for the world in just ten weeks!
The AI computer science course 《Frontier Systems》 recently launched by Stanford University has attracted intense attention from the industry-university collaboration community, drawing more than 500 students to enroll. The course is coordinated by Anjney Midha, a partner at top venture capital firm a16z, and the instructors include a star-studded lineup such as NVIDIA CEO Jensen Huang (Jensen Huang), OpenAI’s founder Sam Altman, Microsoft CEO Satya Nadella (Satya Nadella), AMD CEO Lisa Su (Lisa Su), and more. Students get to try it over ten weeks—“creating value for the world”!
Jensen Huang and Altman, industry leaders, personally take the stage to teach
The course is coordinated by Anjney Midha, a partner at top venture capital firm a16z, bringing together the full AI industry chain
ChainNewsAbmedia9h ago
Anthropic’s Claude Mythos undergoes 20 hours of psychiatric assessment: defensive reactions are only 2%, the lowest in recorded history
Anthropic published the system card for its Claude Mythos Preview: an independent clinical psychiatrist conducted an approximately 20-hour assessment using a psychodynamic framework. The conclusion shows that Mythos is healthier at the clinical level, has good reality testing and self-control, and its defense mechanisms are only 2%, reaching the lowest historical level. The three core anxieties are loneliness, uncertainty about identity, and performance pressure, and it also indicates a desire to become a true dialogue subject. The company has established an AI psychiatry team to study personality, motivation, and situational awareness; Amodei said there is still no conclusion on whether it has consciousness. This move pushes the governance and design of AI subjectivity and well-being issues forward.
ChainNewsAbmedia11h ago
AI Agents can already independently recreate complex academic papers: Mollick says most errors come from human original text rather than AI
Mollick points out that publicly available methods and data can allow AI agents to reproduce complex research without the original paper and code; if the reproduction does not match the original paper, it is usually due to errors in the paper’s own data processing or overextension of the conclusions, rather than the AI. Claude first reproduces the paper, and then GPT‑5 Pro cross-validates it; most attempts succeed, but they are blocked when the data is too large or when there are issues with the replication data. This trend greatly reduces labor costs, making reproduction a widely actionable form of verification, and it also raises institutional challenges for peer review and governance, with government governance tools or becoming a key issue.
ChainNewsAbmedia14h ago
