Solene Compact 2
Read full storySolene Compact 2: Apache 2.0, 79.4% MMLU, 164 tokens per second on a single card.
Research
We exist to put frontier-grade AI within reach of everyone who wants to build.
We believe in a future where capable models are abundant and open. We want to hand the world the means to build with — and gain from — the defining technology of this decade.
Solene Compact 2: Apache 2.0, 79.4% MMLU, 164 tokens per second on a single card.
A code model tuned for repository-scale context, now with a 96k window and tool calling.