Anthropic on Thursday launched Claude Opus 4.6, a significant upgrade to its flagship artificial intelligence model. The company asserts the new version demonstrates improved planning capabilities, sustains longer autonomous workflows, and surpasses competitors, including OpenAI’s GPT-5.2, on key enterprise benchmarks. The release arrives during a period of considerable upheaval in the AI industry and global software markets.
The launch followed closely on the heels of OpenAI’s release of its Codex desktop application just three days prior, a direct challenge to Anthropic’s momentum with Claude Code. Simultaneously, the AI sector has experienced a $285 billion rout in software and services stocks, partially attributed to investor concerns that Anthropic’s AI tools could disrupt established enterprise software businesses.
A key advancement in Claude Opus 4.6 is the introduction of a 1 million token context window. This allows the AI to process and reason across a substantially larger amount of information than previous iterations. Anthropic also introduced “agent teams” within Claude Code, a research preview feature enabling multiple AI agents to collaborate on different aspects of a coding project, coordinating their efforts autonomously.
“We’re focused on building the most capable, reliable, and safe AI systems,” an Anthropic spokesperson told VentureBeat. “Opus 4.6 is even better at planning, helping solve the most complex coding tasks. And the new agent teams feature means users can split work across multiple agents — one on the frontend, one on the API, one on the migration — each owning its piece and coordinating directly with the others.”
The release intensifies the competition between Anthropic and OpenAI, currently the two most highly valued private AI companies. OpenAI’s recent desktop application for its Codex AI coding system aims to transform software development from a collaborative process with a single AI assistant into a management role overseeing a team of autonomous workers. AI coding assistants have seen a surge in popularity, with OpenAI reporting over 1 million developers using Codex in the past month.
The timing of Anthropic’s release – just 72 hours after OpenAI’s Codex launch – underscores the rapid pace of development in AI tools. According to a recent Andreessen Horowitz survey, Anthropic has experienced the largest share increase of any frontier lab since May 2025. Currently, 44 percent of enterprises are utilizing Anthropic in production, driven by rapid gains in software development capabilities since late 2024. OpenAI’s desktop launch is a strategic response to Claude Code’s growing traction.
Anthropic claims Opus 4.6 achieves the highest score on Terminal-Bench 2.0, an agentic coding evaluation, and outperforms all other frontier models on Humanity’s Last Exam, a complex, multi-disciplinary reasoning test. On GDPval-AA – a benchmark measuring performance on economically valuable knowledge work tasks in finance, legal, and other domains – Opus 4.6 surpasses OpenAI’s GPT-5.2 by approximately 144 ELO points, achieving a higher score roughly 70% of the time. Internal testing by Anthropic indicates that Opus 4.6 leads or matches competitors across most benchmark categories, demonstrating particular strength in agentic tasks, office work, and novel problem-solving.
The stakes are considerable. Anthropic announced in November that Claude Code had reached $1 billion in run rate revenue just six months after its general availability in May 2025. Major enterprise deployments include Uber, utilizing Claude Code across software engineering, data science, finance, and trust and safety teams; a wall-to-wall deployment across Salesforce’s global engineering organization; tens of thousands of developers at Accenture; and companies spanning industries like Spotify, Rakuten, Snowflake, Novo Nordisk, and Ramp.
This enterprise traction has fueled a significant increase in the company’s valuation. Earlier this month, Anthropic secured a term sheet for a $10 billion funding round at a $350 billion valuation. Bloomberg reported that Anthropic is also pursuing a tender offer allowing employees to sell shares at that valuation, providing liquidity to staff who have seen the company’s worth multiply since its founding in 2021.
One of the most significant technical improvements in Opus 4.6 addresses the issue of “context rot”—the degradation of model performance as conversations or tasks become more extensive. Anthropic reports that Opus 4.6 scores 76% on MRCR v2, a benchmark testing a model’s ability to retrieve information hidden within large volumes of text, compared to just 18.5% for Sonnet 4.5. The model also supports outputs of up to 128,000 tokens, enabling the completion of substantial coding tasks or documents without requiring segmentation into multiple requests.
For developers, Anthropic is introducing several new API features alongside the model: adaptive thinking, allowing Claude to determine when deeper reasoning is beneficial; four effort levels (low, medium, high, max) to control the balance between intelligence, speed, and cost; and context compaction, a beta feature that automatically summarizes older context to facilitate longer-running tasks.
Anthropic, emphasizing its commitment to AI safety, maintains that Opus 4.6 remains aligned with its predecessors despite its enhanced capabilities. On its automated behavior audit, measuring misaligned behaviors like deception and sycophancy, Opus 4.6 exhibited a low rate of problematic responses while also demonstrating the lowest rate of over-refusals – instances where the model avoids answering legitimate queries – of any recent Claude model.
Regarding safety guardrails as Claude becomes more agentic, the Anthropic spokesperson referenced the company’s published framework, emphasizing the importance of safe, reliable, and trustworthy agents. The company has also developed six new cybersecurity probes to detect potentially harmful uses of the model’s enhanced capabilities and is utilizing Opus 4.6 to identify and address vulnerabilities in open-source software as part of its defensive cybersecurity efforts.
The rivalry between Anthropic and OpenAI has extended into consumer marketing. Both companies will feature prominently during the upcoming Super Bowl. Anthropic is airing commercials that critique OpenAI’s decision to introduce advertisements into ChatGPT, with the tagline: “Ads are coming to AI. But not to Claude.” OpenAI CEO Sam Altman responded by calling the ads “funny” but “clearly dishonest,” stating on X that his company would “obviously never run ads in the way Anthropic depicts them” and that “Anthropic wants to control what people do with AI” while offering “an expensive product to rich people.”
This exchange highlights a fundamental strategic difference: OpenAI is monetizing its large free user base through advertising, while Anthropic is primarily focused on enterprise sales and premium subscriptions.
The launch coincides with significant volatility in software stocks. A new AI automation tool from Anthropic triggered a $285 billion rout in stocks across the software, financial services, and asset management sectors on Tuesday as investors sold shares. A Goldman Sachs basket of US software stocks experienced its largest one-day decline since April’s tariff-fueled selloff. The selloff was triggered by Anthropic’s launch of plug-ins for its Claude Cowork agent, automating tasks across legal, sales, marketing, and data analysis.
Despite the market reaction, Nvidia CEO Jensen Huang stated that fears of AI replacing software tools were “illogical,” and JPMorgan’s Mark Murphy described the selloff as an “illogical leap.”
Anthropic is also releasing Claude in PowerPoint in research preview, allowing users to create presentations using the same AI capabilities as Claude’s document and spreadsheet work. This integration places Claude directly within a core Microsoft product, despite Microsoft’s 27% stake in OpenAI. The Anthropic spokesperson described this as participating in the Office ecosystem and providing users with choice.
Data from a16z’s recent enterprise AI survey indicates that while OpenAI remains the most widely used AI provider, with approximately 77% of surveyed companies using it in production as of January 2026, Anthropic’s adoption is rapidly increasing – from near-zero in March 2024 to approximately 40% in production by January 2026. The survey also shows that 75% of Anthropic’s enterprise customers are using it in production, with 89% either testing or in production, exceeding OpenAI’s 46% and 73% rates, respectively.
Enterprise spending on AI continues to accelerate, reaching an average of $7 million in 2025, up 180% from $2.5 million in 2024, with projections of $11.6 million in 2026 – a 65% year-over-year increase.
Opus 4.6 is available immediately on claude.ai, the Claude API, and major cloud platforms. Developers can access it via claude-opus-4-6 through the API. Pricing remains at $5 per million input tokens and $25 per million output tokens, with premium pricing of $10/$37.50 for prompts exceeding 200,000 tokens using the 1 million token context window. Anthropic recommends adjusting the effort parameter to medium for simpler tasks to optimize cost, and latency.
