# Julia 1 decision model by Supersonic Labs > Julia 1 is a 144M multilingual decision encoder with CPU inference. Read its typed questions, native option limits and separate CPU results. Updated 2026-11-12 (Australia/Adelaide) ## Quick answers ### What is Julia 1 by Supersonic Labs? [Julia 1](): A 144.3M-parameter multilingual encoder adapted from mmBERT-small to score supplied options. [Julia 1 facts]() [Source]() ### Who makes Julia 1? [Julia 1](): Supersonic Labs [Julia 1 facts]() [Source]() ### Are Julia 1's weights open? [Julia 1](): Open weights and packaged Python inference code; private training pipeline Licence: Apache-2.0 declared in the card; no separate package licence file [Julia 1 facts]() [Source]() By Supersonic Labs · Published weights and Python runtime; September catch-up: 2026-09-25 · OPEN A 144.3M-parameter multilingual encoder adapted from mmBERT-small to score supplied options. It accepts named Choice, Noul and ordered Score questions, with a Python runtime for CPU or CUDA. This September release was found through an overnight card edit. ## Facts - Size: 144,292,870 stored parameters; FP32 weights - Inputs: Text state and 2 to 20 options per native question - Context: 8,192 combined tokens supported by the runtime; historical accuracy tests used 1,024 - Access: Open weights and packaged Python inference code; private training pipeline - Licence: Apache-2.0 declared in the card; no separate package licence file ## Claims and sources - **VENDOR-REPORTED**: The maker reports 1,463/2,000 typed questions correct on H200 BF16. A separate CPU FP32 reproduction gets 426/600 Choice, 542/800 Score and 483/600 Noul. [Jev]() figures in the card are supplied references, not newly matched calls; an 8,192-token smoke test does not establish accuracy at that length. [source]() ## Links - [Source: Card, runtime and weights]() - [Source: CPU results]() - [Source: Checkpoint provenance]() ## Related projects No related projects are listed in the registry.