Report: DeepSeek and Moonshot trained new models heavily on GPT-6 Astra outputs
mark_k · x · 2026-10-29
According to The Information, DeepSeek and Moonshot trained their newest models — DeepSeek V4.1-Flash and Kimi K3.1 Pro — heavily on GPT-6 Astra outputs, reportedly switching to Astra after Anthropic added protections against distillation. Neither company has commented; K3.1 Pro's open weights are due Nov 4.
Related event: Report: Moonshot and DeepSeek Distilled OpenAI's Models to Train New AI(3 posts)→
More from Models
- Persona prompt pins models at ~35% on SIMMER benchmark regardless of Claude or DeepSeek — amplifiedamp · 2026-10-29
- OpenAI Terra 6.1 in practice: cheap model plus tiered follow-up beats what benchmarks miss — IamNotGorbachev · 2026-10-29
- qwen-code PR Sizes All Qwen4 Models at 1M-Token Context Window — haider1 · 2026-10-29
- Developer flags Claude weekly limit showing 2% used with zero activity after reset — umeshonai · 2026-10-29
- 'Persona neurons' discourse overhyped, but tiny perturbations breaking alignment is telling — aiamblichus · 2026-10-29
- LocalLLaMA users draft a Qwen4 wishlist: 512K context and plainer answers — perelmanych · 2026-10-29