DeepSeek V4-Pro Tests Reveal Reasoning and Control Issues

Tests of DeepSeek-V4-Pro reveal that its max reasoning setting underperforms compared to high and v4-flash modes, with persistent stop control issues unaffected by prompt changes.

2026-08-16 ~ 2026-08-16 · 2 related posts

• DeepSeek V4-Pro hands-on: Is max reasoning worse than high? Can Harness save it? — karminski3 · 2026-08-16
• DeepSeek V4-Pro's stop-control issue persists across scaffolds; 100M token test shows prompts don't fix it — karminski3 · 2026-08-16