Cogito 32B
by Deep Cogito · cogito family
32B
parameters
text-generation code-generation reasoning math multilingual
Cogito 32B is the mid-size variant from Deep Cogito, built on a Qwen 32B base with enhanced reasoning and multilingual capabilities. It delivers strong performance across mathematical reasoning, code generation, and complex problem-solving tasks. With its 131K context window and competitive benchmark scores, Cogito 32B offers an excellent balance of capability and efficiency for users who need more power than the 8B variant but want to stay within consumer GPU memory limits.
Quick Start with Ollama
ollama run 32b-q4_K_M | Creator | Deep Cogito |
| Parameters | 32B |
| Architecture | transformer-decoder |
| Context | 128K tokens |
| Released | Mar 24, 2025 |
| License | MIT |
| Ollama | cogito:32b |
Quantization Options
| Format | File Size | VRAM Required | Quality | Ollama Tag |
|---|---|---|---|---|
| Q4_K_M rec | 18.5 GB | 21.5 GB | | 32b-q4_K_M |
| Q8_0 | 34 GB | 37 GB | | 32b-q8_0 |
| F16 | 64 GB | 68 GB | | 32b-fp16 |
Compatible Hardware
Q4_K_M requires 21.5 GB VRAM
Benchmark Scores
83.0
mmlu