cogito:3b

Upstream reference: Cogito v1 preview 3B

Publisher: Deep Cogito

Deep Cogito documents this Llama-3.2-derived, instruction-tuned text model with direct-response and explicit thinking modes.

Publisher-described context, separate from Sigma5C observations and recorded catalog statistics.

Published upstream specifications

Upstream base

Meta Llama 3.2 3B Source

Input and output

Text input; text output Source

Published context

128k tokens, using publisher notation Source

Upstream license label

Llama 3.2 Source

Publisher-described uses

• Coding, STEM and instruction-following tasks named in the model card. Source

Reference limitations

• This Llama-derived checkpoint’s license does not establish licenses for Cogito variants derived from other upstream families. Source

A recorded tag and an upstream card association do not verify the deployed checkpoint, quantization, configured context or current hosted availability.

Publisher sources

• Cogito v1 preview 3B model card — Deep Cogito. Checked 2026-11-02.

Links

Recorded references do not establish downloadable weights, licensing or current availability.

• Deep Cogito — official site
• Ollama
• Hugging Face
• Official model announcement / documentation

Where cogito:3b comes from

Named in 4 observation pages

• 3B-7B local models — all 53 · V1
• How models scale under concurrent slots · V1
• NVIDIA GTX 1070 — 344 models profiled · V1
• Sub-3B local models — all 34 · V1