Vendor · Decision models
Mapika has 6 models in the directory: 6 decision models.
In the directory: 6 decision models. 6 of them with open weights that can run locally.
Every model there ever was. Models with a profile lead to every benchmark, price and source.
| Decider 0.8B | n/a | Open source | n/a | 38 |
| Decider 2B | n/a | Open source | n/a | 52 |
| Decider 2B Vision | n/a | Open source | n/a | 36 |
| Decider 35B A3B | n/a | Open source | n/a | 34 |
| Decider 35B A3B NVFP4 | n/a | Open source | n/a | below 1 |
| Decider 4B | n/a | Open source | n/a | 43 |
Decision model · n/a
Decider 0.8B is the smallest text checkpoint in the Decider family. It reads text and typed questions and returns a probability distribution over 2 to 255 supplied options. With 0.8 billion parameters, a 32,000-token context, and local execution, it targets routing, yes-or-no questions, and short states.
Decision model · n/a
Decider 2B occupies the compact middle of the text-based Decider family. The 1.9-billion-parameter checkpoint processes typed questions over a state and distributes probability across 2 to 255 options. Its context window covers 32,000 tokens. The readout happens in one pass, with the answer schema defined before the call.
Decision model · n/a
Decider 2B Vision adds images to the Decider readout. You provide an image, text, and typed questions, and the 1.9-billion-parameter checkpoint returns probabilities over your supplied options. Its context window holds 32,000 tokens.
Decision model · n/a
Decider 35B A3B is the large MoE checkpoint in the Decider family. It contains 34.7 billion parameters, with roughly 3 billion active per token. Over a 32,000-token context, it answers typed text questions in one pass and returns probabilities for your options. Sparse activation still leaves it with the memory footprint of a large local model.
Decision model · n/a
Decider 35B A3B NVFP4 is the quantized form of Decider 35B A3B. The readout stays the same. Attention projections and expert components are quantized to NVFP4; delta-net convolutions and input projections, routers, gates, embeddings, and the LM head remain in bf16. Published memory use therefore drops from 65 GB for the bf16 checkpoint to 19.6 GB. The checkpoint targets vLLM, TensorRT-LLM, and SGLang on Blackwell GPUs.
Decision model · n/a
Decider 4B is the dense 4.2-billion-parameter checkpoint for text decisions. It keeps the same 32,000-token interface as the smaller variants and returns Choice for 2 to 255 options, Score for 2 to 10 levels, and Noul as a yes-probability. Its focus is knowledge and reasoning questions within a fixed answer schema.
By release date. Grey: retired models.
The directory lists 6 Decider models from Mapika: 6 decision models. It counts every model Mapika ever released, including retired ones. 6 are currently available, the newest being Decider 0.8B, Decider 2B and Decider 2B Vision.
Yes, 6 of the 6 models have open weights. Each model profile lists how much memory it needs.
Data as of Nov 1, 2026