BREAKINGOpenAI launches GPT-5 • Apple Vision Pro 2 confirmed for September • NVIDIA Blackwell Ultra ships to cloud providers
Breaking
OpenAI GPT-5 launches with 87% human expert benchmark performanceZepto files DRHP for ₹5B IPO on BSE and NSE — Q3 listing likelyNVIDIA Blackwell Ultra begins shipping to AWS, Azure, Google CloudIndia UPI crosses 20 billion monthly transactions for first timeApple Vision Pro 2 confirmed for September 2026 — starting at $2,499Peak XV Partners closes $2.85B fund, largest India-focused VC fund everTesla FSD 13.0 rolls out to Cybertruck — 847 miles between interventionsSamsung Galaxy S26 Ultra to feature under-display front cameraOpenAI GPT-5 launches with 87% human expert benchmark performanceZepto files DRHP for ₹5B IPO on BSE and NSE — Q3 listing likelyNVIDIA Blackwell Ultra begins shipping to AWS, Azure, Google CloudIndia UPI crosses 20 billion monthly transactions for first timeApple Vision Pro 2 confirmed for September 2026 — starting at $2,499Peak XV Partners closes $2.85B fund, largest India-focused VC fund everTesla FSD 13.0 rolls out to Cybertruck — 847 miles between interventionsSamsung Galaxy S26 Ultra to feature under-display front camera

Anthropic Releases Claude 4 Opus with 2 Million Token Context Window — Sets New AI Benchmark

Anthropic's Claude 4 Opus sets a new benchmark with a 2 million token context window, enabling it to process entire codebases or research libraries in a single conversation.

AS
Aditya Sharma
| 3 May 20265d ago
Share
Anthropic Releases Claude 4 Opus with 2 Million Token Context Window — Sets New AI Benchmark

Anthropic has released Claude 4 Opus, its most capable AI model to date, featuring an unprecedented 2 million token context window.

Context Window Breakthrough

The 2 million token context window — equivalent to approximately 1,500 books or 50,000 pages of text — enables use cases that were previously impossible.

Performance Benchmarks

Claude 4 Opus achieves 96.3% accuracy on GPQA (Graduate-Level Google-Proof Q&A).

Coding Capabilities

Claude 4 Opus demonstrates remarkable coding abilities, solving 87.4% of problems in SWE-bench.