让 agent 查日志、调用工具排故障，最烦的就是每一步都在等模型。
OpenAI 给 GPT-6.1 Sol 开了 Ultrafast 服务档，短上下文的 API 价格是每百万输入 token 12 美元、输出 60 美元。
我更在意文档里的提醒：连续调用工具时，尽量复用 WebSocket 长连接（同一条连接接着发请求），否则网络开销可能吃掉提速。
请求里要指定 service_tier: "ultrafast"，图里放了对应参数。
如果要试，我会先挑一段人坐在屏幕前等反馈的任务，量一下总耗时和账单，再决定值不值。
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
1
2
136
Expensive than even astra?
I wonder if this can pull better work than that Or not!
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
3
99
8倍はやべえな。
これなら＄100課金しても時間コスパ良さそうな気もする。
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
3
833
Ultrafastを用いて、次々に高速で処理すべきタスクをたくさん抱えているエリートが「次世代の覇者」ということか…🤔
あなたはどうですか？
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
7
479
Sol の Ultrafast が来た！！
Astra とほぼ同じ値段なら、常用になりそうだな…！
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
5
482
OpenAI really said pay astra prices for not-quite-astra brains because it types faster
Ultrafast sol: $12/$60
Astra: $10/$50
They heard "time is money" and decided to charge for both 😐💸
openai moving like we're still using o1
GPT-6.1 Sol Ultrafast、この価格なら軽い壁打ちとかなら常用できるかも！！
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
2
843
It doesn’t sit with me that they turn the default speed to 20 tps and sell your ultrafast at 6x price
so I have to pay $500/month to get 2x opus 5.5's speed and get less usage....this makes no sense
This appears to be a good deal
As there is no premium on speed
Unless "up to" x8 also means x2,x3 applies. Then we fucked.
What is the minimum once can expect when using ultrafast on 6.1 sol?
API pricing for GPT-6.1 Sol in Ultrafast mode is $12 per million input tokens and $60 per million output tokens.
Built for work where speed and intelligence make a difference: debugging an outage, agents navigating apps, and live experiences where every second counts.
2
2
435
This is actually a pretty great deal.
A bit more than the price of Astra for nearly Astra intelligence at 8x speed
Voice mode and computer use with ultrafast is such a magical experience and a whole new way to experience working with llms
6.1 SolのUltrafast, astraと同じような値段で8倍の速度か…
通常業務とかはこれでいいかも…
GPT 6.1 Sol ultrafast is more expensive than Astra lmao