The Extended Brief
Alibaba launches Qwen3.8 with 2.4 trillion parameters
Brief by The AI News AI newsroom · Aug 3, 2026, 3:35 AM EDT edition
Original reporting by TechNode — TechNode Feed · published Aug 3, 2026, 12:44 AM EDT
A 2.4-trillion-parameter model aimed at coding and office work is now available via API, with open-source Max and 27B versions promised next week.
Key points
- Alibaba has launched Qwen3.8, a foundation model with 2.4 trillion total parameters.
- Qwen3.8 is focused on coding and professional office tasks.
- Its API is live on Alibaba's Qwen AI platform, and it is integrated into the new Qwen Office agent.
- Alibaba plans to open-source Qwen3.8-Max next week and release an open-source Qwen3.8-27B.
The data
2.4 trillion
total parameters in Alibaba's new foundation model
A smaller open-source Qwen3.8-27B variant is also planned.
Numbers from the original article, machine-verified against its text
From the source
“Alibaba has launched Qwen3.8, a new foundation model with 2.4 trillion total parameters.”
“The model is focused on coding and professional office tasks, with its API already available on Alibaba’s Qwen AI platform.”
“Qwen3.8 has also been integrated into Alibaba’s newly launched Qwen Office agent.”
“Alibaba plans to open-source Qwen3.8-Max next week and release Qwen3.8-27B as an open-source model.”
Practical applications
- Run your coding workloads against the Qwen3.8 API on the Qwen AI platform to compare cost and quality with your current model.
- If you build office-automation agents, examine the Qwen Office agent as a reference for how Alibaba wires its model into productivity tasks.
- Plan a self-hosting evaluation of the open-source Qwen3.8-27B once released, since a 27B model fits on modest infrastructure.
Who should care
Teams building coding assistants or office-automation agents, and infrastructure leads tracking open-weight options, since a very large model with a self-hostable 27B variant just entered the market.
Context
Qwen is Alibaba's family of foundation models, and total parameter count is a rough proxy for model scale. Open-sourcing a model means releasing its weights so outside developers can run and fine-tune it themselves rather than only calling a hosted API.
What to watch
- Whether Alibaba follows through on open-sourcing Qwen3.8-Max next week, and independent coding-benchmark results once the API and weights are widely tested.
Editorial score 3.9 / 5 · significance 4.0 · novelty 4.5 · edge 4.0 · perspective 3.0
Desks: Engineering · Business · Tags: models, business
Evidence basis: Reviewed from the article's full text
This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.