Together AI has released GLM-5.3-Flash, also known as Ox Alpha, positioning it as a top choice for intelligence per dollar. The model reportedly achieves 63% performance at a cost of $0.24 per task. This release offers a competitive option compared to DeepSeek-V4 Flash 0731 and GPT-5.6 Luna, which have different performance and cost metrics. AI
IMPACT Offers a new cost-effective option for AI inference, potentially driving down operational costs for users.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on X — Together (inference / OSS) →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →