65,000 real API calls: GPT-6.1 Sol is the slowest of 15 models in actual usage
RexDouglass · x · 2026-10-05
After users complained GPT-6.1 Sol feels sluggish, one ran an agent over 65,000 calls from his coding tools since August, comparing answers of similar length and measuring wait time from request to full response.
In his real-world data, GPT-6.1 Sol ranked dead last among the 15 models he used — even Fast mode only matched regular GPT-6 Sol. Not a benchmark, just actual usage.
More from Models
- Sentdex: GLM 5.3 Flash is "the best model on the planet," runnable locally — Sentdex · 2026-10-05
- Dev slams OpenAI Codex as buggy and slow, says Claude Code pulls ahead — lyraxana · 2026-10-05
- Burkov: for 95% of users a chatbot is just a search engine plus paraphrased copyright violation — burkov · 2026-10-05
- Codex computer use + Qwen 3.6 35B: 'never felt computer use be so fast' — TheZachMueller · 2026-10-05
- GPT-6 Astra tops 4 Design Arena leaderboards, leads 3D design by wide margin — BorisMPower · 2026-10-05
- Codex User Races to Burn 97% of Usage Quota in 4.5 Hours — Angaisb_ · 2026-10-05