AI model
Qwen3.6-27B is a 27-billion-parameter open-weight language model from Alibaba's Qwen family, used in benchmarks, local inference and research.
Sakana AI and UC Berkeley's MASS method reached a 69.9% win rate against its Qwen3.6-27B base model after two rounds of self-training. The same model ranks its own runs, and its evaluator sometimes passes flawed steps into the next training set.
Reality
MLCommons' MLPerf Inference v6.1 adds tests that time a whole RAG pipeline and a 1,007-turn coding-agent replay. Those scores help a buyer size hardware only where the buyer's own pipeline spends its time the way the four-model reference does.
Reality
UW and Meta researchers report 59.4% on BrowseComp-Plus for a model that edits its own context, against 53.4% for Codex-style summarisation. The edited file is thrown away when the task ends, so memory that lasts across tasks is still the builder's job.
Reality
One 8,192-token session on a 27B model holds 512 MB of key-value cache, or 64 KB for every token generated. How many of those sessions fit in free VRAM sets serving concurrency, and paging decides the waste.
Reality
Anthropic screened Claude for one property, verbalizability, and found two more in the same representations. Neel Nanda reproduced the structure in open weights, and the outside commentators Anthropic invited disagree about what it is.
Reality
NVIDIA's submission runs the same Qwen3.6-27B as the llama.cpp reference on the same Jetson board and finishes 6.4x sooner. Most of the gap comes from prompt tokens the runtime never has to prefill.
Reality
JetBrains has taken the assembly work out of running a coding agent offline. What it could not take out is the hardware, and that is now the thing deciding who adopts.
Reality
Muse Glimmer ships as Apache 2.0 weights sized for a 24GB card. Muse Spark 1.2 stays on Muse Code and the Meta Model API. Plan capacity for two tiers, not one.
Perspective Coverage
6 publishersReality