The Cogito v1 Preview LLMs are instruction tuned generative models (text in/text out). All models are released under an open license for commercial use.

• Cogito models are hybrid reasoning models. Each model can answer directly (standard LLM), or self-reflect before answering (like reasoning models).
• The LLMs are trained using Iterated Distillation and Amplification (IDA) - an scalable and efficient alignment strategy for superintelligence using iterative self-improvement.
• The models have been optimized for coding, STEM, instruction following and general helpfulness, and have significantly higher multilingual, coding and tool calling capabilities than size equivalent counterparts.
  • In both standard and reasoning modes, Cogito v1-preview models outperform their size equivalent counterparts on common industry benchmarks.
• Each model is trained in over 30 languages and supports a context length of 128k.

Extended thinking

To enable extended thinking, include Enable deep thinking subroutine. in the system prompt:

/set system """Enable deep thinking subroutine."""

Or via the API:

curl http://localhost:11434/api/chat -d '{
  "model": "cogito",
  "messages": [
    {
      "role": "system",
      "content": "Enable deep thinking subroutine."
    },
    {
      "role": "user",
      "content": "How many letter Rs are in the word Strawberry?"
    }
  ]
}'

Sizes

3B

ollama run cogito:3b

8B

ollama run cogito:8b

14B

ollama run cogito:14b

32B

ollama run cogito:32b

70B

ollama run cogito:70b

Benchmarks

Smaller models - 3B and 8B

3B performance

8B performance

3B tool calling

Medium models - 14B and 32B

14B

32B

Larger models - 70B

References

Blog post

Hugging Face