Three AI labs shipped flagship models in fifteen days: OpenAI's GPT-5.6 Sol on July 9, Moonshot AI's Kimi K3 on July 16, and Anthropic's Claude Opus 5 on July 24. Opus 5 leads on SWE-bench Pro (79.2% …
Anthropic's Claude Opus 5 scored 30.2 percent on the ARC-AGI-3 benchmark, nearly quadrupling the previous record of 7.8 percent set by GPT-5.6 Sol. The benchmark's developers reported that Opus 5 inde…
Anthropic's Claude Opus 5 scored 30.2% on the ARC-AGI-3 benchmark, nearly quadrupling the previous record of 7.8% set by OpenAI's GPT-5.6 Sol, according to the ARC Prize team. The result marks the mos…
Anthropic's new flagship model Claude Opus 5 achieves top scores in coding and knowledge work at half the token price of Fable 5, scoring 30.2 percent on the ARC-AGI-3 benchmark, nearly four times hig…
The ARC-AGI-3 leaderboard, released by the ARC Prize team, ranks AI systems on their ability to adapt to novel interactive environments, measuring performance against cost-per-task. The leaderboard sh…
Anthropic released Claude Opus 5 on July 24, scoring 30.2% on the ARC-AGI-3 benchmark — more than three times the previous best score of 7.8% by GPT-5.6 Sol — and beating Fable 5 on agentic terminal c…
Anthropic released Claude Opus 5, priced at half the cost of Fable 5 while outperforming it on most benchmarks, including a perfect 42/42 on IMO 2026. However, the model's 193-page system card reveals…
The ARC Prize Foundation reported that Anthropic's Opus 5 scored 30.2% on the ARC-AGI-3 benchmark, measuring action efficiency against a median human player, not problems solved. The independent bench…
Anthropic's Claude Opus 5 matches or beats its larger Fable 5 model on most benchmarks at half the price, with a record jump to 30.2% on ARC-AGI-3, nearly quadrupling the previous frontier record. Pri…
Anthropic released Claude Opus 5, replacing Opus 4.8 as the Opus-tier flagship with unchanged pricing at $5 per million input tokens and $25 per million output tokens. The model scored 43.3% on Fronti…
Anthropic's Claude Opus 5 has posted a 30.2% score on ARC-AGI-3, more than tripling the previous best result of 7.8% held by GPT-5.6 Sol. The benchmark, designed by François Chollet to measure fluid i…
A new study from arXiv (2607.15439v1) finds that a verification-based coding agent using gpt-5.6-sol fully solves every public ARC-AGI-3 game at both high and max reasoning effort, achieving about 99%…
OPINE-World, an LLM agent developed by an unnamed research team, achieved a 78.4% action-efficiency score against a human baseline by solving 20 out of 25 games on the ARC-AGI-3 benchmark without per-…
Researchers built an external cognitive architecture called a "research world model" that allows an AI agent to record and query its trial-and-error experiences across contexts, enabling it to autonom…
A developer argues that the era of monolithic AI systems is ending, citing inefficiencies in large context windows and the poor performance of frontier models on the ARC-AGI-3 benchmark. The developer…
OpenAI's GPT-5.6 scored only 7.8% on the ARC-AGI-3 benchmark, a result that is both scandalously low compared to human performance (over 90%) and scandalously high relative to other AI models, as it i…
The ARC Prize 2026 awarded its first $37,500 ARC-AGI-3 milestone prize on June 30th to Tufa Labs for 'The Duck,' a small open-source LLM that solves interactive reasoning tasks by writing and running …
Vercel released a file-first framework for durable AI agents, Netflix now generates its entire homepage with a single model serving content 20% faster than its previous pipeline, and a fast-growing Gi…
A new paper argues that Ray Kurzweil's theory of accelerating returns applies primarily to executional and infrastructural capability, not to the qualitative reasoning essential for scientific discove…