DeepSeek routing drama: switch to v4.1 flash slammed, rollback to Pro called flip-flopping
teortaxesTex · x · 2026-09-11
A viral post mocks DeepSeek's no-win routing situation: switching Pro to v4.1 flash drew complaints that better eval scores don't mean better real-world output, while rolling back to the original Pro after user feedback got it accused of breaking its word — with some speculating the new flash model underperforms Pro. The author argues vendors should stick to their decisions, joking that Anthropic embodies this trait well.
Related event: DeepSeek routing backlash: silent Flash switch rolled back(2 posts)→
More from Models
- ToneBench eval: GPT-6.1 Terra ranks #23 at writing — cheaper, not better, than GPT-5.6 Terra — OnlyProggingForFun · 2026-10-28
- Cursor user complains it silently switches to Grok, which fails basic tasks — kylegawley · 2026-10-28
- Musk says Grok Bot keeps improving, calls Vault feature a game changer — beffjezos · 2026-10-28
- Claude refuses to gather legitimately public information, sparking guardrail backlash — mkheck · 2026-10-28
- DesignArena launches Code to Video Arena: models direct 15-45s films purely in code — yshan2u · 2026-10-28
- ChatGPT's 'yes, but with an important clarification' often precedes pure nonsense — Crixusgannicus · 2026-10-28