Development

Chinese open models take the majority of token use

development2026-06-20

Claim

Qwen, DeepSeek, Kimi, GLM and MiniMax now command the majority of token use across OpenRouter's top ten, up from under 2% in late 2024, with GLM-5.2 one-shotting a deliberately AI-resistant take-home to beat Claude Opus 4.8 on readable, maintainable code.

Reported metric: <2% to majority in 18 months

Reporting date: 2026-06-20 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to June 20, 2026 (theinnermostloop.substack.com)

A former DeepMind VP called it the first open model good enough to be a daily driver — a win partly down to the Fable rollback.

Knowledge graphopen in browser →

Reported in

Welcome to June 20, 2026

Actor

Evidence for

Supersedes

A 744B open model is shrunk from 1.51TB to 217GB to run locally

Referenced by

Cite or correct this record

Citation formatCitation text

Chinese open models take the majority of token use
Innermost Loop KB — secondary record
Reporting date: 2026-06-20
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-06-20-open-chinese-models-take-the-majority
Claim: Qwen, DeepSeek, Kimi, GLM and MiniMax now command the majority of token use across OpenRouter's top ten, up from under 2% in late 2024, with GLM-5.2 one-shotting a deliberately AI-resistant take-home to beat Claude Opus 4.8 on readable, maintainable code.
Attributed sources:
- Welcome to June 20, 2026 — https://theinnermostloop.substack.com/p/welcome-to-june-20-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.