Users suspect DeepSeek V4 quality drop due to routing changes
l33thax0r_ · reddit · 2026-08-28
A developer observed that DeepSeek V4 Flash on OpenCode Go seems to have degraded recently, often missing the point despite long responses. Coinciding with official pricing changes, the user suspects model routing policy shifts rather than model updates. They plan to use ZenMux to compare different routing paths to verify this hypothesis.
More from Models
- Planned Rerun of 3D Game Benchmark for Grok 4.6 and GLM 5.3 Flash — kevinkern · 2026-08-28
- Local LLMs Still Lagging Behind Frontier Models — sdmat123 · 2026-08-28
- Own a frontier AI model running locally in just 5 hours — MaziyarPanahi · 2026-08-28
- Qwen3.8-Flash-Next Released: A Free AI Rivaling Billion-Dollar Giants — Two Minute Papers · 2026-08-28
- DeepSeek V4 Flash Pricing Matches GPT OSS 20B, Sparking Cost Discussion — ChrizBogota · 2026-08-28
- PhoneLLM Alpha 1 Trends on Hugging Face: Voice Agent Model with MoE — pipecat-ai · 2026-08-28