Compact open models and data-constrained training approaches are achieving competitive results on specific tasks, from sub-4B models to BabyLM-style budgets. Efficiency is becoming a primary design lever for cost, deployment and on-device use.
In the BabyLM 2026 Strict-Small, Qiushi Engine gained 0.23 Overall points across nine metrics without increasing scale, but by reorganizing experience around contextual dependencies. For teams evaluating self-hosted LLMs, the relevant signal is the d...
Qiushi Engine ran an end-to-end autonomous research program on BabyLM 2026 Strict-Small, using 10 million corpus words and 100 million cumulative word presentations. Three stages linked frontier advancement, principle discovery, and principle-guided ...
A new framework tackles heterogeneous vocabularies in collaborative speculative decoding. X-CoSD limits distribution exchange to the common region between the two models; the extended variant moves resampling to the server and leaves the device with ...
OpenBMB has released MiniCPM5-2B, a 2-billion-parameter open weights model scoring 15 on the Artificial Analysis Intelligence Index v4.2, the highest among open models up to 4B. For self-hosted deployments, the small size and open weights lower the p...
A test on 25 Nutrition5k meals compares seven multimodal LLMs on calorie estimation under a 20% error threshold. Open-weights Muse Spark 1.3 hits 48%, while Qwen 3.8 27B stops at 16%. The ranking doesn't track model size—a useful signal for anyone ev...
A llama.cpp pull request adds support for Spark-X2.5, two compact 4B and 1.7B parameter LLMs with a native context window of up to 1M tokens and a hybrid sliding-window attention architecture. The move lowers the barrier for local and self-hosted inf...
An experiment puts a 90M-parameter conversational model on a 2004 Sony console. Inference runs at about 0.5–0.6 tokens per second, with replies taking one to three minutes. More a technical provocation than a useful tool, but it says a lot about real...
