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
OpenClaw Releases v2026.4.20 with Kimi K2.6 as Default Model
Gate News message, April 21 — OpenClaw, an open-source AI Agent platform, released v2026.4.20 on April 20, introducing significant updates to default model switching, plugin contract handling, and gateway storage optimization.
The release switches the default model for Moonshot's built-in
GateNews17m ago
OpenAI Launches ChatGPT Images 2.0 Model with Enhanced Visual Task Capabilities
Gate News message, April 21 — OpenAI unveiled ChatGPT Images 2.0 on April 22, significantly enhancing complex visual task processing capabilities. The model improves instruction understanding, object placement and relationship expression, and high-density
GateNews25m ago
SpaceX Agrees to Acquire AI Startup Cursor for Over $50 Billion
Gate News message, April 21 — According to The New York Times, SpaceX has agreed to acquire Cursor, a fast-growing artificial intelligence startup focused on code-writing software, for over $50 billion. The deal comes as SpaceX prepares for what could be one of the largest initial public offerings i
GateNews33m ago
Unauthorized Users Access Anthropic's Mythos AI Model, Company Confirms Investigation
Gate News message, April 22 — A small group of unauthorized users accessed Anthropic's new Mythos AI model on the same day the company announced plans to release it to select organizations for testing, Bloomberg News reported. The users, part of a private online forum, gained access to Mythos and
GateNews42m ago
Silicon Valley’s ‘monitoring the situation’ MTS meme becomes a 24/7 news machine delivered by a16z
a16z is backing "Monitoring the Situation," a 24/7 X livestream born from Polymarket meme culture, as tech VCs build their own news-industrial complex.
Summary
Andreessen Horowitz has helped launch "Monitoring the Situation" (MTS), a 24/7 livestream show on X, leaning into crypto-prediction mar
Cryptonews4h ago
Google launches Deep Research Max: supports MCP and can access enterprise private data
According to a Google DeepMind official blog announcement, Google launched a new generation of autonomous research agents, Deep Research and Deep Research Max, on April 21, 2026. They are built on Gemini 3.1 Pro, as the official release following the preview version provided through the Interactions API in December 2025. Both agents are now available in the paid tiers of the Gemini API in the form of a public preview, and new startups and enterprise users on Google Cloud will be able to connect gradually.
The two variants are positioned differently: interactive vs asynchronous depth
Google categorizes the two agents by use case: Deep Research
ChainNewsAbmedia6h ago
