The written, source-linked counterpart to this video.
Published edition3 stories
01Updates
Z.AI plans financing for its next generation of models
Z dot A I has published plans for a new share placement and a proposed convertible bond issue to fund its next generation of models. Its September thirteenth filing sets the bond principal at twenty point one four billion renminbi.
The Placing and the issue of the Bonds are independent of each other and are not inter-conditional.
The filing does not establish that all the money has arrived, or that a new model is already available.
Watch next
Watch for the next Z dot A I filing confirming whether the share placement and bond issue have closed.
Intern S two now has a public model release for scientific and general agent tasks. The official repository was created on September thirteenth and lists the weights under the Apache two point zero license.
Intern-S2-397B can be deployed using any of the following LLM inference frameworks:.
The reported benchmark results come from the model team. They are not independent tests, and open weights do not mean that the hardware needed to run the model is free.
Watch next
Check the Intern S two deployment documentation and test the model on your own task. The documentation says time series inference currently requires L M Deploy.
LangChain has published how its paid media agent works and linked the open source project. The agent combines advertising data with warehouse results and brings proposed campaign changes into Slack for review.
Use models for judgment and code for consistency.
Teams can inspect that workflow without copying every tool or loading an entire company wiki into each prompt. This is a company case study. Its reported savings are not a promise for another business, and the workflow still needs connected data, account permissions and approval for campaign changes.
Watch next
Review the LangChain repository permissions and verify the data sources and the result of every approved change before expanding access.