The first purpose-built mixture-of-experts model for Data Engineering and Data Science. Expert SQL generation, PySpark authoring, data modeling, and autonomous multi-step orchestration, in a single 9B sparse model that activates only 2.6B parameters per token.
A mixture-of-experts transformer that routes each token to the two experts that matter. You get the knowledge of a larger model at the inference cost of a small one.
Post-trained on curated data engineering and data science tasks, then reinforcement-tuned against executed code.
Generation, optimization, execution-plan reasoning, and dialect translation like Snowflake to BigQuery.
Author and optimize Spark pipelines end to end, with lakehouse and streaming and CDC patterns built in.
Speaks TL fluently and emits declarative Bonacci pipelines you can run as they are.
Autonomous multi-step orchestration from design to deploy, reasoning across the whole pipeline.
Calls tools, retrieves with RAG-aware generation, and uses extended thinking for hard problems.
Data modeling, quality and observability, and security-aware generation across the lakehouse.
Broad language understanding, code, and reasoning foundations.
High-quality educational and technical web content.
Curated DE and DS pairs: SQL, PySpark, ThinkingLanguage, pipeline blueprints, and schemas.
Reward-shaped on SQL correctness, query-plan quality, and PySpark execution.
An OpenAI-compatible endpoint with vLLM, or load it in Transformers. bfloat16 weights, 32K context.
# OpenAI-compatible endpoint $ vllm serve thinkingdbx/bonacci-moe-9b \ --dtype bfloat16 \ --max-model-len 32768 \ --tensor-parallel-size 2 \ --gpu-memory-utilization 0.90
from transformers import AutoModelForCausalLM, AutoTokenizer import torch model_id = "thinkingdbx/bonacci-moe-9b" tok = AutoTokenizer.from_pretrained(model_id) model = AutoModelForCausalLM.from_pretrained( model_id, torch_dtype=torch.bfloat16)
Model id and weights publish at launch.
It already powers parts of Bonacci Studio. When the weights publish, we will email everyone who asked. No spam, one message at launch.