AI Weekly Malaysia

Reflection announced Beam, a text-only 501B-total / 23B-active MoE for coding, agentic, and scientific work, with full Apache 2.0 weights due this month. It cites 23.8T pretraining tokens, RL on ~10,500 GB300s, and claimed 80.9 SWE-bench Verified plus 3–4x the inference efficiency of GLM 5.2. Independent reads place it around GLM-5.2 and below DSv4 Flash on some benchmarks, while estimating ~12% BF16 MFU and a DeepSeek V3-like iso-FLOP architecture.

Why: Builders evaluating coding agents should plan to test Beam when the Apache 2.0 weights land this month: the claimed 80.9 SWE-bench and 3–4x efficiency vs GLM 5.2 are attractive, but the text says it trails GLM 5.3, Kimi K3, Qwen 3.8 Max, and DeepSeek V4.1 Flash, so it is likely a cheaper open option rather than a clear upgrade. No Malaysia-specific policy, funding, infrastructure, or provider detail appears in the text.

Mistral released a preview of Mistral Large 4, a 1-trillion-parameter model with 49 billion active parameters, trained on Mistral's own cluster of 3,800 NVIDIA Grace Blackwell GPUs and available now only through their API. The preview exposes just two reasoning levels, "none" and "high", and Mistral promises open weights at the end of this month. On Artificial Analysis it scores 38, behind DeepSeek 4.1 Flash (a 552B model), a large jump from Mistral Large 3's score of 9 in December, though Simon Willison describes it as roughly six months behind the frontier.

Why: If you self-host or care about open weights, this is an API-only preview today, so any evaluation has to wait for the end-of-month weight release — don't plan deployments on the API tier unless you're fine with a hosted-only dependency. The two-level reasoning switch (none vs high) is unusually coarse: the "high" pelican test used fewer output tokens (2,717) than "none" (3,275), so you can't assume "high" costs more output tokens when budgeting. Compared with DeepSeek 4.1 Flash scoring higher at 552B, the practical question is whether a 1T/49B-active MoE gives you enough quality per dollar to justify swapping out your current model.

This is a short Simon Willison blog post (6 Oct 2026) linking to his Hacker News comment on Mistral Large 4. The post quotes an HN commenter, wren6991, joking that 'the benchmark is saturated' and that frontier models are 'tested with an armadillo in fishnet tights jaywalking on Mars' — Willison then runs that exact prompt through claude-opus-5.5, gpt-6.1-sol, gemini-3.8-flash, and mistral/mistral-large-4 at default reasoning levels using his llm CLI, and links an SVG renderer to compare the outputs. No benchmarks, pricing, context window, or capability claims about Mistral Large 4 are given.

Why: There is almost nothing here to act on: no scores, no pricing, no API details, no changelog for Mistral Large 4 — just a four-model SVG side-by-side at default reasoning settings. The one usable takeaway is a workflow, not a fact: if you're evaluating a new model release, Willison's pattern (one odd prompt, run across models with `llm -m <model>` at default settings, outputs rendered and compared visually) is a cheap sanity check you can copy before trusting any leaderboard. Do not make a model-selection decision for your product based on this item alone; the text does not support one.