Dicta-LM 3.0: Advancing The Frontier of Hebrew Sovereign LLMs

Dicta-LM 3.0 is a powerful open-weight collection of LLMs, trained on extensive corpora of Hebrew and English texts. The models are available for download and for unlimited use. The models set a new SOTA for their weight-class for Hebrew, both as base models and chat models.

This is our flagship model, a 24-billion-parameter reasoning model, with full precision (BF16), originally initialized from Mistral-Small-3.1-24B-Base-2503.

This model is a reasoning chat model, which means that before responding to any given message from the user, the model first thinks out the right way to respond in a designated thinking block.


πŸš€ Try it out here: chat.dicta.org.il


For full details of this model please read our release blog post or the technical report.

You can view and access the full collection of base/instruct unquantized/quantized versions of DictaLM 3.0 here.

Instruction format

In order to leverage instruction fine-tuning, your prompt should be rendered using the chat template specified for this model. Most libraries deal with this automatically, so you can just let them do it.

Usage

We recommend using vLLM, but you can use Transformers as well:

Transformers


from transformers import pipeline

generator = pipeline('text-generation', model="dicta-il/DictaLM-3.0-24B-Thinking")

messages = [
    {"role": "user", "content": "איזה Χ¨Χ•Χ˜Χ‘ אהוב Χ’ΧœΧ™Χš?"},
    {"role": "assistant", "content": "Χ˜Χ•Χ‘, אני Χ“Χ™ ΧžΧ—Χ‘Χ‘ Χ›ΧžΧ” Χ˜Χ™Χ€Χ•Χͺ ΧžΧ™Χ₯ ΧœΧ™ΧžΧ•ΧŸ Χ‘Χ—Χ•Χ˜ Χ˜Χ¨Χ™. Χ–Χ” ΧžΧ•Χ‘Χ™Χ£ Χ‘Χ“Χ™Χ•Χ§ אΧͺ Χ”Χ›ΧžΧ•Χͺ Χ”Χ Χ›Χ•Χ Χ” של טגם Χ—ΧžΧ¦ΧžΧ₯ ΧœΧ›Χœ ΧžΧ” שאני ΧžΧ‘Χ©Χœ Χ‘ΧžΧ˜Χ‘Χ—!"},
    {"role": "user", "content": "האם Χ™Χ© לך מΧͺכונים ΧœΧžΧ™Χ•Χ Χ–?"}
]

print(generator(messages)[0]['generated_text'][-1]) # just print the last message
# {'role': 'assistant', 'content': 'Χ”ΧžΧ©Χͺמש Χ©Χ•ΧΧœ: "האם Χ™Χ© לך מΧͺכונים ΧœΧžΧ™Χ•Χ Χ–?" Χ’ΧœΧ™Χ Χ• ΧœΧ”Χ’Χ™Χ‘ גם מΧͺΧ›Χ•ΧŸ ΧœΧžΧ™Χ•Χ Χ–. Χ‘Χ€Χ§Χ• מΧͺΧ›Χ•ΧŸ Χ§ΧœΧΧ‘Χ™, ΧΧ•ΧœΧ™ גם וריאציוΧͺ (ΧžΧ™Χ•Χ Χ– Χ€ΧœΧ™ΧΧ•, ΧžΧ™Χ•Χ Χ– Χ˜Χ‘Χ’Χ•Χ Χ™, ΧžΧ™Χ•Χ Χ– Χ—Χ¨Χ™Χ£, Χ•Χ›Χ•\'). Χ‘Χ€Χ§Χ• הוראוΧͺ Χ©ΧœΧ‘ אחר Χ©ΧœΧ‘. Χ¦Χ¨Χ™Χš ΧœΧ”Χ™Χ•Χͺ Χ‘Χ˜Χ•ΧŸ Χ™Χ“Χ™Χ“Χ•ΧͺΧ™.</think>Χ‘Χ˜Χ—! ΧžΧ™Χ•Χ Χ– Χ‘Χ™ΧͺΧ™ הוא Χ‘ΧΧžΧͺ Χ€Χ©Χ•Χ˜ ΧœΧ”Χ›Χ Χ” Χ•ΧžΧ’Χ Χ™Χ§ ΧœΧ›Χ Χ©ΧœΧ™Χ˜Χ” ΧžΧœΧΧ” גל Χ”ΧžΧ¨Χ›Χ™Χ‘Χ™Χ, Χ”Χ˜Χ’Χ Χ•Χ”ΧžΧ¨Χ§Χ. ΧœΧ”ΧœΧŸ Χ”ΧžΧͺΧ›Χ•ΧŸ Χ”Χ§ΧœΧΧ‘Χ™, Χ‘ΧͺΧ•Χ‘Χ€Χͺ Χ›ΧžΧ” וריאציוΧͺ ΧžΧ”Χ Χ•Χͺ Χ©ΧͺΧ•Χ›ΧœΧ• ΧœΧ”ΧͺΧ Χ‘Χ•Χͺ Χ‘Χ”ΧŸ.\n\n---\n\n## 1️⃣ ΧžΧ™Χ•Χ Χ– Χ§ΧœΧΧ‘Χ™ (ΧžΧ™Χ•Χ Χ– Χ‘Χ‘Χ™Χ‘Χ™)\n\n| ΧžΧ¨Χ›Χ™Χ‘ | Χ›ΧžΧ•Χͺ |\n|------------|----------|\n| Χ—ΧœΧžΧ•ΧŸ Χ‘Χ™Χ¦Χ” Χ’Χ“Χ•Χœ (או 2 Χ—ΧœΧžΧ•Χ Χ™Χ בינוניים) | 1 |\n| Χ—Χ¨Χ“Χœ Χ“Χ™Χ–\'Χ•ΧŸ | 1 Χ›Χ€Χ™Χͺ |\n| ΧžΧ™Χ₯ ΧœΧ™ΧžΧ•ΧŸ Χ˜Χ¨Χ™ (או Χ—'}

vLLM

vllm serve dicta-il/DictaLM-3.0-24B-Thinking --enable-auto-tool-choice --tool-call-parser hermes --reasoning_parser deepseek_r1

And then you can access it via the openai library:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="sk-no-key-required"
)

response = client.chat.completions.create(
    model="dicta-il/DictaLM-3.0-24B-Thinking",
    messages=[
        {"role": "user", "content": "Hello, how are you?"}
    ],
)

print(response.choices[0].message.content)

The reasoning traces should be available in the response structure in the designated fild.

The model supports tool-calling, enabling integration with external tools and APIs. For example how to use the tool calling, see the vLLM documentation.

Citation

If you use this model, please cite:

@article{Shmidman2025DictaLM3,
  title={{Dicta-LM 3.0: Advancing The Frontier of Hebrew Sovereign LLMs}},
  author={Shaltiel Shmidman and Avi Shmidman and Amir DN Cohen and Moshe Koppel},
  year={2025},
  publisher={{DICTA / Jerusalem, Israel}},
  note={https://www.dicta.org.il/publications/DictaLM_3_0___Techincal_Report.pdf}
}
Downloads last month
1,753
GGUF
Model size
24B params
Architecture
llama
Hardware compatibility
Log In to view the estimation

1-bit

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for unsloth/DictaLM-3.0-24B-Thinking-GGUF

Quantized
(1)
this model

Collection including unsloth/DictaLM-3.0-24B-Thinking-GGUF