DeepSeek Accused of Grey Testing for High Scores, Opus 5 Output Quality Questioned
teortaxesTex · x · 2026-08-16
Users suggest DeepSeek may be using grey tests to achieve high scores, with DSH minimal mode converting DeepSeek credits into Opus 5-level outputs, but actual SVG quality falls short. The discussion also touches on other providers and pricing.
More from Models
- 8 hours testing Pangram: anti-AI-detection tricks all failed; only self-typed drafts pass — PawelHuryn · 2026-08-16
- Uncensored Qwen 27B Model Released as GGUF Quantization — BLUECOW009 · 2026-08-16
- Grok 4.6 beats GPT-5.6 Sol on coding agent efficiency, 35% lower cost — rohanpaul_ai · 2026-08-16
- Uncensored vision-enabled Qwen3.6-27B finetune trends on Hugging Face with GGUF release — HauhauCS · 2026-08-16
- Adjusting prediction effort busts cache; watermarking remains — banteg · 2026-08-16
- Anthropic details how Claude’s new watermarks work: mechanism and resistance to editing — TechCrunch AI · 2026-08-16