Article Complete
AI NewsContextual Overkill

Anthropic to Release Claude 3.7 with a 500k Context Window: Reports

This signifies a massive jump from the 200k context window.

Anthropic to Release Claude 3.7 with a 500k Context Window: Reports

Key takeaways

  • Anthropic will launch Claude 3.7 with a groundbreaking 500k context window, up from 200k.
  • This expansion allows users to input significantly larger content, enhancing various applications.
  • User excitement grows, particularly for coding use cases, highlighting the previous limitations of the 200k window.
  • Competitors like Google's Gemini 2.0 Flash model lead with a 1 million context window.
  • Claude 3.7 Sonnet previously demonstrated superior accuracy compared to OpenAI's o3-mini in model evaluations.

Takeaways drafted with AI from the article text.

Anthropic plans to release a new version of Claude 3.7 Sonnet with a 500k context window, a significant jump from the current 200k capacity. TestingCatalog spotted evidence of the same in feature flags. 

The expansion will allow users to input significantly larger content in the prompt. The report also added that it isn’t clear if the expansion will be available to all plans. Several users on X have expressed their excitement in anticipation of the update. “500k would be amazing for coding use cases,” said a user on X. 

Another said the news is “music” to their ears. “The 200K context window is the biggest shortcoming in Claude Code right now,” they added.

Similarly, OpenAI’s o3-mini and the o1 models also support a 200,000 context window. 

Meanwhile, the GPT-4.5 research preview model offers a 128,000 context window. However, Google’s Gemini 2.0 Flash model has a 1 million token context window, along with the newly released Gemini 2.5 Pro. Google also announced that a 2 million context window is coming soon. 

Last month, the company released the Claude 3.7 Sonnet, a hybrid model with extended thinking capabilities. It scored 62.3% accuracy on the SWE-bench while OpenAI’s o3-mini (high) scored a 49.3%. Artificial Analysis, a platform that independently analyses AI models, called it the best non-reasoning model for coding. 

Alongside Claude 3.7 Sonnet, the company also released Claude Code, an agentic coding tool that operates directly within the terminal. It is capable of fixing bugs across a code base, resolving merge conflicts, creating commits and pull requests, and answering questions about the architecture and logic.

The company also recently added a web search feature to Claude, a long-missing feature now available in preview for paid users in the US. According to their blog, free users and more regions will gain access soon. 

Users can enable web search in the Claude web app’s settings, allowing the chatbot to pull information from the internet when needed. Currently, the feature works only with Claude 3.7 Sonnet.

Published

Corrections: AIM corrects errors of fact and notes the change on the story. Spotted one? Tell the editors. Read our editorial standards.

More in this story

AI (Artificial Intelligence) · 5 stories
  1. Applications For Google’s Startups Accelerator India Program Now Open
  2. Baidu To Launch ChatGPT Rival
  3. Tencent Releases Hunyuan Turbo S to Rival DeepSeek as Competition Heats up in China
  4. You are here: Anthropic to Release Claude 3.7 with a 500k Context Window: Reports
  5. OpenAI’s Coding Provider Turing Raises $111 Million at $2.2 Billion Valuation 
About the author

Contributor

All stories
Belamy · Daily

The day’s AI news, before your first meeting.