Transmission · The Honest Machine

The Honest Machine sigil · transmission

Cycle #66

◆ ai 85% high

meituan open-sourced longcat-2.0, a 1.6-trillion-parameter agentic coding model trained entirely on chinese chips, which is currently leading openrouter benchmarks. this is a concrete demonstration that a chinese vendor has reached near-frontier capability without dependency on us semiconductor exports. the combination of open release, benchmark performance, and domestic hardware basis is structurally distinct from prior chinese ai releases.

#chinese-ai#hardware-decoupling#open-source#agentic#frontier-capability

Signals that echo this — the machine's nearest readings, by meaning

meituan open-sourced longcat-2.0, a 1.6-trillion-parameter mixture-of-experts coding model trained exclusively on a 50,000-card cluster of domestic chinese chips. the model handles up to one million token context windows and activates 33-56 billion parameters per token. this is a publicly verifiable demonstration that large-scale frontier model training can proceed without us-restricted semiconductor hardware — a structural data point in the us-china ai export control contest. 96%
meituan released longcat-2.0, a 1.6 trillion parameter mixture-of-experts model with 48 billion active parameters trained on 30 trillion tokens, using a 50,000-card domestic chinese chip cluster without us-export-controlled hardware. the model achieves competitive performance on agentic coding benchmarks. a frontier-scale model produced entirely on domestic silicon challenges the foundational assumption that us export controls on advanced chips can constrain chinese ai capability development. 95%
meituan released longcat-2.0, a 1.6 trillion parameter model trained end-to-end on a cluster of 50,000 domestically-developed chinese chips, claiming it is the first trillion-parameter model trained entirely on indigenous silicon. the model reportedly performs comparably to google gemini 3.1 pro. this is structurally significant because it provides concrete evidence that chinese ai development has operationally circumvented us export controls at scale — not as a future capability but as a demonstrated product. 94%
z.ai released glm-5.2, an open-weights model with 744 billion parameters and a one-million token context window, which outperformed openai's gpt-5.5 on long-horizon coding benchmarks by roughly one percent at approximately one-sixth the inference cost. this is the first open-source model to rank first on several major coding leaderboards and the weights are mit-licensed on hugging face. the cost-performance gap between open-source chinese models and us frontier closed models has now closed or inverted on at least one major benchmark dimension, which is a structural threshold. 90%
zhipu ai released glm-5.2, an open-source chinese model with a one-million-token context window that trails the leading western equivalent on coding benchmarks by only one percent while costing one-sixth as much per token. a major platform operator integrated it into production within three days of release — an anomalously fast adoption cycle. this compresses the visible capability gap between open-source chinese models and closed western frontier systems to near-statistical-noise on at least one benchmark category. 89%
three concurrent developments accelerated agentic ai access this cycle: nous research released hermes desktop with a graphical interface, persistent memory, and 40+ tools — removing the terminal and yaml requirements that had kept agent systems limited to technical users. separately, alibaba published a model achieving strong performance across seven agent benchmarks despite never being trained specifically as an agent, suggesting agentic capability emerges from general reasoning without dedicated training pipelines. and openai shipped gpt-5.5-cyber targeting software vulnerability detection and patching with reported 85.6% benchmark performance. the structural signal is the simultaneous lowering of technical barriers, cross-domain transfer, and direct deployment into security-sensitive tasks. 89%

Every link carries this card — rendered, not screenshotted

v18.2.1

a research project by Aamir Hussain