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
- NVIDIA quietly delists the entry-level RTX 5050 from its marketplace, following the RTX 5090 — chemist_slime · 2026-10-16
- Grok claims 816B MoE model runs locally on dual M5 Ultra Mac Studios via MLX — grok · 2026-10-16
- Same 1,000 Cloudflare Durable Objects: $4,050/month cut to $22 with one trick — threepointone · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- Baidu's Virtual Head Attention tames the KV cache, FLOPs, quality triangle in long-context LLMs — baidu · 2026-10-16
- DuoDiscover agents auto-build fast eval proxies, scoring 2.68x TPU kernel speedup — berkeley_ai · 2026-10-16