Ring-2.6-1T is a 1T-parameter-scale thinking model with 63B active parameters, built for real-world agent workflows that require both strong capability and operational efficiency. $0 per million input tokens, $0 per million output tokens. 262,144 token context window, maximum output of 65,536 tokens.
#inclusionAI: Ring-2.6-1T
29 messages · Page 1 of 1 (latest)
Where benchmark
doesn't look like they released one yet
@deft ledge think u were interested in this
ðŸ˜
wonder if its any good
ling had a terrible bench iirc
why do all their models end in ing
ling ring
bing chilling
Where is Ping?
Well it seems quite smart and token-efficient, latency is currently amazing too
And text-to-voice model, Sing
10T100AB model - King
Model specialised in uncensored cursed web search - Bing
It was the Ling one I had wanted to see and it was unfortunately a steaming heap
follows instructions very precisely too, very good
not super great at anticipating needs for UI but quite competent at coding otherwise
this model will probably be eaten by the claws
is that another way of saying hot shit
lmao
mord is very articulate with his words 