Programmer Ditches Top Models: Cost per Task Varies 20x, Speed Is the New Agentic AI Bottleneck
ievkz · reddit · 2026-09-18
A programmer found he no longer needs the smartest, priciest models. Key metric: cost per completed task — GPT-5.6 Luna (max) costs $0.18 vs $3.26 for GPT-6 Astra (max), nearly 20x apart (AA Intelligence Index: 38 vs 53). Switching to Luna's cheapest tier ended his rate-limit frustrations on a $20/month plan. The new bottleneck is speed: most models on DeepInfra/OpenRouter run 25-50 tokens/s (5-15 min per Codex task), while DeepSeek-V4.1-Flash via its own API hits 300 tokens/s with a higher intelligence index of 40. His take: the race for smarter models is ending, the agentic race is now about speed, and Chinese models currently lead — while OpenAI's upcoming data centers may push thousands of tokens/s. Context-window race is over too; its growth is what spawned coding agents.
Related event: Dev ditches top model, finds nearly 20x cost gap per task(2 posts)→
More from coding & agent
- Butterfly traces multi-agent failures with graphs, 62.5% accuracy vs 53.8% for log prompting — JeremyCMorgan · 2026-11-04
- Free course covers the 4 hard problems of production generative media agents — Al_Grigor · 2026-11-04
- Quota reset night: Codex back to 100% while Claude shows just 10% left — rudrank · 2026-11-04
- Self-improving software is inevitable: closing the agent-observability loop — bibryam · 2026-11-04
- Building a 7-GPU 132GB VRAM local inference rig: consolidate on EPYC or split across two systems? — infectiousstupidity · 2026-11-04
- Founder of 50-person, 12-country team says he talks to AI agents more than humans — Scobleizer · 2026-11-04