OpenAI has released GPT-6.2 Astra and GPT-6.2 Sol, with GPT-6.2 Astra seeing a significant price reduction in cached input costs. This price cut is particularly beneficial for agent-based workloads that frequently resend large context prefixes. While the headline input and output prices are also lower for GPT-6.2 Astra, the cache reduction offers the most substantial savings for specific use cases. The new models also feature updated knowledge cutoffs and some changes to API migration rules. AI
IMPACT The significant reduction in cached input costs for GPT-6.2 Astra could lower operational expenses for agent-based AI applications.
RANK_REASON Frontier-lab model release with system card and pricing details. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
- AIHubMix
- Anthropic
- GPT-6.1 Sol
- GPT-6.2 Astra
- GPT -6 Astra
- .openai
- DataStudios
- DevTk
- Hostinger
- TokenLean
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →