AI Weekly Malaysia

Aleph Alpha released Kolibri on 3 October 2026, an open-weight German/English mixture-of-experts LLM with 78.1B total parameters but only 3.46B active per token, under Apache 2.0 for the weights and config files (training code and methods stay proprietary). It was trained from scratch on ~24 trillion tokens — over a fifth German — on 768 NVIDIA B200 GPUs using infrastructure in Germany and Finland, with a 262,144-token native context (tested to 1,048,576), four reasoning levels, tool calling, a 18 June 2026 knowledge cutoff, and about 78 GB of FP8 weights. Aleph Alpha frames it as 'sovereign': built under European/German law with no foreign control, so customers get full deployment freedom and 'compliance as an inherited property', and it has signed the EU's GPAI Code of Practice. The 409-point Hacker News thread drew only 11 comments.

Why: The ~78 GB FP8 footprint means Kolibri can plausibly run on a single 80 GB accelerator rather than a cluster, which is the concrete difference between self-hosting and paying per-token to a US API. If you sell into the EU, handle data that cannot leave a client's building, or need tool-calling agents with a 262k context window, this is a deployable alternative — but the 'scores above every compared model of its size in both languages' claim comes from Aleph Alpha's own evaluation, so benchmark it yourself before committing. For Malaysian and SEA builders, the relevant lesson is the packaging: weights + license + no-foreign-control deployment story as a compliance argument, which is a template local sovereign-model efforts can copy.

Parley is a federated, decentralised chat server that speaks plain IRC: you run an instance for your own domain and people talk to anyone as user@domain from irssi or any existing IRC client. The repo (git.mills.io/prologic/parley) shows 192 commits, 5 tags/releases, 4 open issues, 1 watcher, 2 stars and 1 fork, and it hit 210 points with 101 comments on Hacker News. The latest visible commit is a chat change about pushing a mention only to the mentioned user.

Why: If you run a community or internal chat and want to avoid a client-rollout project, Parley's pitch is that members keep using whatever IRC client they already have and just get an identity like user@domain — no new app to install or train people on. But the repo signals are tiny (2 stars, 1 fork, 4 open issues, 5 releases), so treat this as something to spin up in an afternoon and evaluate, not something to migrate a live community onto this week. There is no Malaysia-specific detail in this item; the only local angle is that a Malaysian dev group could self-host an instance on its own domain instead of renting a Discord or Slack workspace.

A firm rented four Nvidia H200 GPUs at $13,200 per month to independently test DeepSeek's claim of being '80x cheaper', and the rental alone reportedly doubled what the firm was already paying for Claude. The same write-up notes that security flaws forced the team to keep their code offline during the test. The article body itself did not load in the supplied text, so no benchmark results, token throughput, or final verdict are available here.

Why: The only concrete numbers we have are the cost side: $13,200/month for four H200s versus an existing Claude bill that this doubled, plus a security constraint that kept code off the network entirely. If you are weighing self-hosted or rented-GPU inference against API spend, this is a reminder that the comparison is rental + ops + isolation overhead, not just per-token price — and that the '80x cheaper' figure is still unverified here. Because no results are in the text, don't cite this as evidence either way yet; wait for the actual measurements.