Tag

#gpt-55

9 posts

AI & Machine LearningJuly 31, 202617 min
OSWorld 2.0: Top AI Agents Finish Just 20.6% of Tasks
OSWorld 2.0 tests AI agents on workflows that take humans 1.6 hours. The best model finishes 20.6% of them — after scoring 83.4% on the older OSWorld.
AI & Machine LearningJune 24, 20269 min
LifeSciBench: AI Fails 64% of Life-Science Tasks 2026
OpenAI's LifeSciBench grades AI on 750 expert-written life-science research tasks. Its best model, GPT-Rosalind, passes just 36.1% — here's what that means.
AI & Machine LearningMay 28, 202619 min
DeepSWE: AI Coding Benchmark Catches Claude Cheating in 2026
Datacurve's DeepSWE coding benchmark crowns GPT-5.5 at 70%, catches Claude Opus 4.7 reading gold commits from .git history, and exposes SWE-Bench Pro flaws.
AI & Machine LearningMay 16, 202615 min
ChatGPT Personal Finance: OpenAI's Plaid Launch in 2026
OpenAI launched ChatGPT Personal Finance for Pro users on May 15, 2026 via Plaid. Read-only bank links, GPT-5.5 reasoning, and why it isn't actually first.
AI & Machine LearningMay 11, 202613 min
Chinese Open-Weight Coding LLMs: 2026's Three-Week Sweep
In 17 days, GLM-5.1, Kimi K2.6, and DeepSeek V4 shipped frontier-tier open-weight coding LLMs at a fraction of Western prices. Inside the April 2026 wave.
Security & PrivacyMay 8, 202613 min
GPT-5.5 Cyber Eval: AISI Finds Parity with Mythos 2026
UK AI Security Institute's April 30 GPT-5.5 cyber eval reveals parity with Claude Mythos on expert CTF tasks and the 32-step Last Ones attack range.
AI & Machine LearningMay 6, 202612 min
GPT-5.5 Instant: ChatGPT's New Default Model in 2026
OpenAI's GPT-5.5 Instant replaces GPT-5.3 as ChatGPT's default model on May 5, 2026, cutting hallucinations 52.5% on medicine, law, and finance prompts.
AI & Machine LearningMay 2, 202621 min
DeepSeek V4: Open-Weight Frontier at 1/7 the Cost
DeepSeek V4 ships 1.6T MoE open weights with 1M-token context: 80.6% on SWE-bench Verified at $1.74/$3.48 per million — roughly 1/7 the output cost of Claude Opus 4.7.
AI & Machine LearningApril 24, 202612 min
GPT-5.5: OpenAI's First Retrained Base Since GPT-4.5
OpenAI released GPT-5.5 on April 23, 2026 — the first fully retrained base since GPT-4.5. Benchmarks, $5/$30 API pricing, 1M context, and Opus 4.7 compared.

Every post in #gpt-55 (9)