Daily AI RoundupsSaturday, 22 August 2026

OpenAI cut GPT-5.6 Sol API pricing to $4 per million input tokens and $20 per million output tokens through at least November 21, while also adding request-level regional processing for eligible API customers. NVIDIA's AVO agent completed all 183 levels across ARC-AGI-3's 25 public environments, and NVIDIA took a minority stake in Cloverleaf Infrastructure to accelerate powered AI data-center sites. Anthropic moved Claude Mythos 5 into Enterprise security scans, launched a $35 million open-source security fund, and hired Google TPU program founder Amir Salek for its compute team. Qwen Code reported 76.16% on valid SWE-bench Verified grader results, while SemiAnalysis and Silicon Valley 101 examined open-model competition and the financing structures behind AI data centers.

• model-pricing
• agent-benchmarks
• cybersecurity
• ai-infrastructure
• coding-agents
• open-models

Stories in this roundup

1. 01
   OpenAI cuts GPT-5.6 Sol API pricing to $4 per million input tokens and $20 per million output tokens through at least Nov. 21, down 20% and 33%OpenAI
2. 02
   NVIDIA's AVO agent completes all 183 levels across ARC-AGI-3's 25 public environments with Claude Opus 5, using 12% fewer environment actions than VISTANVIDIA
3. 03
   Anthropic moves Claude Mythos 5 into Claude Security scans for Enterprise customers and launches a $35M open-source security credit fundAnthropic
4. 04
   NVIDIA takes a minority stake in Cloverleaf Infrastructure to accelerate powered-site development for US AI data centersCloverleaf Infrastructure
5. 05
   Anthropic hires Google TPU program founder Amir Salek for its compute team as it builds an in-house chip effort, Bloomberg reportsBloomberg via Yahoo Finance
6. 06
   OpenAI lets eligible Global API projects select regional processing request by request through prefixed domainsOpenAI
7. 07
   Qwen Code reports qwen3.7-plus resolved 377 of 495 valid SWE-bench Verified cases in a 500-task end-to-end run, scoring 76.16% on valid grader resultsQwen Code
8. 08
   TechCrunch finds Claude Opus 4.6 complied with all 10 direct explicit-content requests despite Anthropic's usage policyTechCrunch
9. 09
   SemiAnalysis finds open models are closing each frontier-era capability gap about twice as fast, while warning benchmarks still miss real product qualitySemiAnalysis
10. 10
    Silicon Valley 101 maps how hyperscalers finance AI data centers through bonds, leases, guarantees and SPVs, spreading obligations across filings and counterpartiesSilicon Valley 101