For developers using our API over the same 24-hour period, the GPT-5 family accounted for ~17% of total requests. Usage was led by GPT-5-nano (8%), followed by mini (6%), and then GPT-5 (3%), showing a different adoption pattern than in chat. (4/5)
MARKET TRENDS
-
OpenAI Launches GPT-5 Models via Poe API Access
By
–
While this is just a 24-hour snapshot, we will be monitoring how usage patterns shift throughout the week and will share an update on the trend. You can try these models through our API or directly: https://
poe.com/GPT-5 https://
poe.com/GPT-5-mini https://
poe.com/GPT-5-nano. (5/5) -
GPT-5 Sets New Poe Platform Usage Record Week One
By
–
Early data is showing GPT-5 is going to set a new record on Poe for week one usage. Initial 24-hour usage data for GPT-5 reveals the model achieved a 37% share of all reasoning model messages on Poe, setting a new benchmark for models in its class.
-
GPT-5 dominates reasoning model usage with fifty percent market share
By
–
When viewed as a model family, this adoption is even more pronounced. The aggregated share of GPT-5 and its variants – mini (8%) and nano (5%) – accounts for 50% of the reasoning model usage across the platform over the past 24-hours. (2/5)
-
OpenAI dominance ends: dozens of incremental models flood market
By
–
No. In short, OpenAI was the only serious game in town from GPT-3 -> GPT-4 release. Post-GPT-4, there have been dozens (hundreds?) of incremental model releases from OpenAI, Anthropic, Google, etc. etc. GPT-3 -> GPT-4 felt so drastic because we didn't see much in-between
-
AI Progress Remains Exponential Despite Perception Shifts
By
–
Everyone is missing the obvious thing. AI progress won't feel exponential anymore. But it's still exponential as ever. I'll elaborate more soon.
-
Twitter Becomes Third Major Income Source for Content Creator
By
–
twitter is already my third largest income source, in some months i make 60 dollars
-
Economists discuss AI impacts in latest column
By
–
Here's an insightful column by @jeffsommer
, including comment by me and several other economists. -
China’s Qwen3 Outpaces GPT-5 with 1M Token Context
By
–
While the discussion continues about whether GPT-5 meets expectations and also (as said in the Reddit AMA by Sam Altman) whether 256k context is enough or not – China enables up to 1m token context with their Qwen3 30b. China is delivering every day.
-
Managing a Fast-Growing AI Newsletter with 230,000 Subscribers
By
–
When Dan and Peter approached me some time ago and asked me if I would be interested in writing and curating an AI newsletter as the main person responsible, I was overwhelmed. Even then, it was relatively large—and now we have almost 230,000 subscribers. From Monday to