OpenAI's Ultrafast tier drops to ~65 tok/s, breaking paid priority routing plans
thepriceisright__ · reddit · 2026-10-17
A Reddit developer reports that over the past week, throughput on the gpt-6.1-sol Ultrafast endpoint has dropped to roughly 65 tok/s — the same speed as the cheaper Fast tier, making the two service tiers effectively indistinguishable.
The poster was building a feature that routes requests to Ultrafast as paid "high priority" processing, but says the plan is unworkable if OpenAI can't keep speeds consistent or even distinct across tiers. A throughput comparison screenshot is attached, and the thread asks whether others are seeing the same degradation.
More from Models
- DeepSeek flash 1017 called out for still acting "very much like Claude" — teortaxesTex · 2026-10-17
- Anthropic says distillation defenses cut harvested data 90-99%, names Zhipu GLM-5.3 — bibryam · 2026-10-17
- JevBench Adds 34 Decision Models in 4 Days as kettei-fat Takes #1 — airesearch12 · 2026-10-17
- Diffusion Gemma promises 4x faster text generation; may supplant autoregressive LLMs — NandoDF · 2026-10-17
- Perplexity reveals pplx-decider: reranker with zero reranking supervision, private MTEB-style eval — bo_wangbo · 2026-10-17
- Anthropic's code-review skill helps Opus 5.5 fix 5.5 more planted bugs at lower cost — PawelHuryn · 2026-10-17