Unverified50% confidenceFactExact time
Large language models (LLMs) get their name from their parameter scale, with modern mainstream models typically having parameter counts ranging from tens of billions to trillions.
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCore large model R&D personnel typically earn annual salaries of over one million RMB.65% similarUnverifiedThe vocabulary of mainstream large models typically contains 30,000 to 150,000 Tokens.64% similarUnverifiedThe K3 large model released by Kimi has a parameter scale of nearly 3 trillion (2.8 trillion).62% similarUnverifiedThe financial threshold required to train a frontier large model has climbed to the hundreds of millions of dollars.61% similarUnverifiedAlibaba open-sourced the weights of a trillion-scale flagship model for the first time, with a total of 2.4 trillion parameters, approximately 95 billion parameters activated per token, using a MoE sparse architecture.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/927586API
curl https://kongchang.com/api/v1/knowledge/claims/927586MCP
get_claim(id=927586)