Open weight models are impressive, but Deepseek continues to impress
Summary
Open weight models provide cost benefits, but Deepseek is advancing AI technology with architectural improvements that may influence the broader field.
Similar Articles
The Unbearable Cheapness of Open Weight Models
The article examines the dramatic cost difference between open-weight models like DeepSeek V4 and closed models from Anthropic and OpenAI, arguing that the latter sustain high prices through artificial scarcity and branding rather than technical superiority.
@FinanceYF5: 阅读文章：https://openrouter.ai/blog/insights/the-open-weight-models-that-matter-june-2026/…
The article highlights the growing importance of open-weight AI models as of June 2026, with DeepSeek V4 Flash emerging as a cost-effective, high-performance model that rivals frontier models like GPT-5.5 for agentic tasks.
@AdinaYakup: DeepSeek V4 Pro's weights are out https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813… - 1.6T / 49B MoE, MIT licens…
DeepSeek released DeepSeek-V4-Pro-0813, an open-weights MoE model (1.6T total / 49B active) with MIT license, 1M context, and enhanced agentic capabilities, outperforming its preview and competing with top proprietary models on benchmarks like Terminal-Bench 2.1.
)
DeepSeek permanently reduced V4 Pro prices by 75%, undercutting leading AI models from OpenAI, Anthropic, and Google, escalating the AI price war.
)
DeepSeek unveiled its new V4-Pro-0813 model, priced at $0.87 per million output tokens, while claiming strong benchmark performance against Anthropic's Opus 4.8 and intensifying the AI price war with OpenAI.