BERLIN — August 13, 2025 — Anthropic has taken a major step forward in its AI capabilities with the launch of a dramatically expanded context window and a new, on-demand memory feature for its Claude AI platform. These innovations mark a significant escalation in the competition among enterprise-focused AI systems from Anthropic, OpenAI, and Google.


1. Claude Sonnet 4: Now with 1 Million Tokens of Context

Anthropic announced that its Sonnet 4 model now supports a context window of up to 1 million tokens—a substantial five-fold increase over the previous 200,000-token limit.Wikipedia+15VentureBeat+15Tom’s Guide+15WIRED+15Anthropic+15Neowin+15

  • Use cases unlocked: Developers can now process entire codebases with over 75,000 lines of code, or analyze dozens of lengthy documents—like legal contracts or research papers—in a single request without losing context.Reddit+6Anthropic+69to5Mac+6
  • Pricing adjustments: To accommodate the increased computational load, prompts exceeding 200K tokens will incur higher costs—$6 per million input tokens and $22.50 per million output tokens, up from the earlier $3 / $15 pricing tiers.Anthropic+2TechCrunch+2
  • Customer highlights: Bolt.new and iGent AI report significant productivity gains, with Bolt.new highlighting improved project-scale coding workflows and iGent’s Maestro agent empowered by the expanded context window.Anthropic+1
  • Competitive context: While Anthropic’s Sonnet 4 is among the first to push toward enterprise-scale context length, other players like OpenAI’s GPT‑4.1 achieved similar capabilities earlier in April.Reddit+8The Verge+8IT Pro+8

2. On-Demand Chat Recall: A Memory Feature with Privacy in Mind

Anthropic also released a new memory tool that enables Claude to retrieve and reference past conversations—but only when explicitly prompted by the user.

  • This differs from ChatGPT’s persistent memory, which automatically builds and updates user profiles.
  • Anthropic’s design places user privacy at the forefront: no background profiling, only manual, user-initiated access to prior chat content.Tom’s Guide+3The Verge+3Tom’s Guide+3

3. Why It Matters: Enterprise Momentum & Competitive Positioning

These upgrades are strategic for Anthropic’s positioning in the enterprise AI market:

FeatureBenefit for UsersCompetitive Impact
1 M-token contextEnables handling entire codebases and large document sets in one requestCloses gap with OpenAI/Google on long-context capabilities
Memory featureSupports continuity in multi-session projects without compromising privacyDifferentiates Claude with a privacy-first design
  • Developer workflows become significantly smoother—no need to chop large tasks into smaller chunks.
  • Data-rich industries like legal, pharma, and software engineering stand to benefit most from these expanded capabilities.
  • Privacy-conscious enterprises may lean toward Claude for its opt-in memory design.

4. Verdict: Claude Strengthens Its Role as a High-Context, Privacy-Aware AI Assistant

Anthropic’s upgrades to Claude Sonnet 4—particularly the 1-million-token context window—dramatically enhance its utility in handling large-scale, complex workflows like full codebase analysis or document consolidation. The memory tool adds usability for sustained, multi-session tasks while maintaining a privacy-first approach.

As Anthropic continues rolling out these features more broadly in the coming weeks, Claude further positions itself as a high-context, enterprise-oriented AI rival to OpenAI’s ChatGPT and Google’s Gemini.

You May Also Like

AI in the Public Sector: Government AI Projects Making an Impact

For those interested in how AI is transforming government services, explore the impactful projects shaping future governance today.

AI Patent Wars: Rising Competition Over AI Innovations and IP

Discover how the fierce AI patent wars are transforming innovation and IP battles worldwide, with implications that could reshape the future of technology.

AI Revolution Drives Layoffs: Inside Atlassian’s Controversial Workforce Reshuffle

The layoff that shocked tech workers On 30 July 2025 Atlassian’s 12,000‑person workforce received…

Anthropic’s December Enterprise Signals: What Procurement, Risk, and Platform Teams Should Notice

By Thorsten Meyer AI December delivered a set of quiet but consequential…