Architecture
data/models.yaml. Every label is auditable
against the model's sources.
Specs
- Architecture
- dense
- Total params
- not disclosed
- Active params
- 104B
- Context window
- not verified
- Attention
- gqa
- Position encoding
- rope
- Post-training
- sft, dpo
- OSI-approved
- no
- Data released
- no
- Training code
- not released
Benchmarks
Each score carries the date it was published; we never infer or interpolate missing scores.
Recommended use cases
- retrieval-augmented generation with citations
- agentic tool-use workflows
- non-commercial research under CC BY-NC
Available quantizations
GGUF llama.cpp's container; the common local format, k-quants from Q2 to Q8.
runs on llama.cpp, Ollama
FP8 8-bit float, frequently a native release on Hopper / Blackwell GPUs.
runs on vLLM, SGLang, TensorRT-LLM
Verified via the Hugging Face model tree ↗. Community quantizations change over time; the families shown are those with published weights at audit time.
Notable innovations
- · Grounded generation with citation templates
- · Multi-step tool use in the base chat template
- · Optimized for 10 business languages
Known limitations
- · CC BY-NC 4.0 forbids commercial use of the open weights; production use requires the Cohere API. source ↗
Lineage
Larger sibling of Command R (35B). Superseded by Command R+ 08-2024 and later Command A releases on the production API.