Model Card for Cohere Labs Command A Translate

Model Summary

Cohere Labs Command A Translate is an open weights research release of a 111 billion parameter model that achieves state-of-the-art performance on translation quality.

Developed by: Cohere and Cohere Labs

• Point of Contact: Cohere For AI: Cohere Labs
• License: CC-BY-NC, requires also adhering to Cohere Lab's Acceptable Use Policy
• Model: command-a-translate-08-2025
• Model Size: 111B
• Context length: 8k input, 8k output

For more details about this model, please check out our blog post and technical report.

Try Cohere Labs Command A Translate

You can try out Cohere Labs Command A Translate before downloading the weights in our hosted Hugging Face Space.

Usage

Please install transformers from the source repository that includes the necessary changes for this model.

# pip install transformers
from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "CohereLabs/command-a-translate-08-2025"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

# Format message with the command-a-translate-08-2025 chat template
messages = [{"role": "user", "content": "Translate everything that follows into Spanish:\n\n 'Enterprises rely on translation for some of their most sensitive and business-critical documents and cannot risk data leakage, compliance violations, or misunderstandings. Mistranslated documents can reduce trust and have strategic implications.'"}]
input_ids = tokenizer.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_tensors="pt"
)

gen_tokens = model.generate(
    input_ids, 
    max_new_tokens=4096, 
    do_sample=True, 
    temperature=0.3,
)

gen_text = tokenizer.decode(gen_tokens[0])
print(gen_text)

You can also use the model directly using transformers pipeline abstraction:

from transformers import AutoTokenizer, pipeline
import torch

model_id = "CohereLabs/command-a-translate-08-2025"

pipe = pipeline(
    "text-generation",
    model=model_id,
    torch_dtype="auto",
    device_map="auto",
)

messages = [
    {"role": "user", "content": "Translate everything that follows into Spanish:\n\n 'Enterprises rely on translation for some of their most sensitive and business-critical documents and cannot risk data leakage, compliance violations, or misunderstandings. Mistranslated documents can reduce trust and have strategic implications.'"},
]

tokenizer = AutoTokenizer.from_pretrained(model_id)

text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
)


outputs = pipe(
    messages,
    max_new_tokens=300,
)
print(outputs[0]["generated_text"][-1])

Model Details

Input: Text only.

Output: Model generates text.

Model Architecture: This is an auto-regressive language model that uses an optimized transformer architecture. After pretraining, this model uses supervised fine-tuning (SFT) and preference training to align model behavior to human preferences for helpfulness and safety. The model features three layers with sliding window attention (window size 4096) and RoPE for efficient local context modeling and relative positional encoding. A fourth layer uses global attention without positional embeddings, enabling unrestricted token interactions across the entire sequence.

Languages covered: The model has been trained on 23 languages: English, French, Spanish, Italian, German, Portuguese, Japanese, Korean, Arabic, Chinese, Russian, Polish, Turkish, Vietnamese, Dutch, Czech, Indonesian, Ukrainian, Romanian, Greek, Hindi, Hebrew, and Persian.

Context Length: Command A Translate supports a context length of 8K input & 8K output length.

Model Card Contact

For errors or additional questions about details in this model card, contact labs@cohere.com.

Terms of Use:

We hope that the release of this model will make community-based research efforts more accessible, by releasing the weights of a highly performant 111 billion parameter model to researchers all over the world. This model is governed by a CC-BY-NC License (Non-Commercial) with an acceptable use addendum, and also requires adhering to Cohere Lab's Acceptable Use Policy. If you are interested in commercial use, please contact Cohere’s Sales team.

Try it now:

You can try Command A Translate in the playground here. You can also use it in our dedicated Hugging Face Space here.

Citation

@inproceedings{kocmi-etal-2025-command,
    title = "Command-A-Translate: Raising the Bar of Machine Translation with Difficulty Filtering",
    author = {Kocmi, Tom  and
      Arkhangorodsky, Arkady  and
      Berard, Alexandre  and
      Blunsom, Phil  and
      Cahyawijaya, Samuel  and
      Dehaze, Th{\'e}o  and
      Fadaee, Marzieh  and
      Frosst, Nicholas  and
      Galle, Matthias  and
      Gomez, Aidan  and
      Govindarajan, Nithya  and
      Ko, Wei-Yin  and
      Kreutzer, Julia  and
      Marchisio, Kelly  and
      {\"U}st{\"u}n, Ahmet  and
      Vincent, Sebastian  and
      Zhang, Ivan},
    editor = "Haddow, Barry  and
      Kocmi, Tom  and
      Koehn, Philipp  and
      Monz, Christof",
    booktitle = "Proceedings of the Tenth Conference on Machine Translation",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.wmt-1.55/",
    doi = "10.18653/v1/2025.wmt-1.55",
    pages = "789--799",
    ISBN = "979-8-89176-341-8",
}