DeepSeek is 90x Cheaper Than Opus, But Opus Still Wins on Every Row
mustafamhus · x · 2026-08-01
Mustafa Suleiman pointed out that while DeepSeek is 90x cheaper than Claude 3 Opus on output (and 36x cheaper on input), Opus still wins every row of DeepSeek's own benchmark table. Furthermore, Opus's advantage widens as tasks get harder.
He concluded that the correct approach to this cost-performance tradeoff is to "route, don't replace"—dynamically routing tasks to models based on complexity rather than blindly replacing capable models with cheaper ones.
More from Infra
- Llama 3.1 405B Hits 5.6k t/s on Cerebras for Select Customers — kimmonismus · 2026-08-01
- MediaTek Expects 400G/448G SerDes IP Ready by H2 Next Year — rwang07 · 2026-08-01
- Google Unveils 8th-Gen TPU with Up to 80% Better Performance-Per-Dollar — tekbog · 2026-08-01
- Self-hosting Kimi K3 takes 2 years to break even, revealing cloud AI economics — Liu_eroteme · 2026-08-01
- Why Kubernetes Is the Wrong Primitive for Diverse AI Agent Workloads — blaizedsouza · 2026-08-01
- Open Source Personal AI Computer Project Enables Local Desktop LLMs — dee_hw · 2026-08-01