DeepSeek V4 Models to Feature Revamped Reasoning Modes, Leak Suggests
teortaxesTex · x · 2026-08-03
According to developer discussions and repository details, DeepSeek's upcoming V4 series is adjusting its reasoning effort modes. The V4-Flash will effectively feature four modes: nothink, low, high, and max, while V4-Pro omits the 'high' tier. The company has also revamped its reasoning reinforcement learning (RL) for the General Availability (GA) release.
Related event: DeepSeek V4 Preview: Flash to Introduce Four-Level Reasoning Effort(2 posts)→
More from Models
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Argon edges out astra as daily driver, but Claude still wins heavy coding work — iruletheworldmo · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- RoboVista benchmark: Gemini-4 Argon hits 81.9% on robotics understanding — berkeley_ai · 2026-10-16
- Dev reports decision model jev beats gpt-6-luna on quality in real training pipeline — genmon · 2026-10-16
- Google won't publish Argon 4 limits in AI Pro—users compare experiences — Consistent-Text-2662 · 2026-10-16