Use DictaLM 3.0 24B Thinking for Free

DictaLM 3.0 24B Thinking is an AI model released 2025 by Dicta, a leading organization in large language model research. Hebrew-specialized LLM with thinking mode — best model for Hebrew language tasks with chain-of-thought reasoning. With 24B and a a large context window context window, DictaLM 3.0 24B Thinking is designed to excel across a range of demanding tasks — from software engineering and mathematical problem-solving to multilingual content generation and long-document analysis.

DictaLM 3.0 24B Thinking uses a dense transformer architecture with 24B of parameters, all of which are active for every input token. Dense architectures provide predictable, consistent behavior across diverse task types — there are no "cold" experts to warm up, and quality is uniform across prompts. Dicta's training pipeline combined large-scale pretraining on a large-scale multilingual pretraining dataset with supervised fine-tuning and alignment techniques (RLHF or DPO), producing DictaLM 3.0 24B Thinking's characteristic balance of capability, helpfulness, and safety.

As of 2026-10-16, DictaLM 3.0 24B Thinking by Dicta demonstrates measurable performance on standard AI evaluation benchmarks. On Hebrew NLU — which measures model performance on specialized evaluation tasks — it achieves 79.3%. These scores place DictaLM 3.0 24B Thinking — at 24B of parameters — in a competitive position within its model class. Note that benchmark scores reflect controlled evaluation conditions; real-world performance varies by task, prompt design, and domain specificity.

The primary strengths of DictaLM 3.0 24B Thinking are: First large Hebrew-specialized AI model with thinking/reasoning mode; 24B parameters dedicated to Hebrew language understanding; Critical for Israeli tech ecosystem and Hebrew-language enterprise AI. These make it particularly well-suited for Hebrew-language enterprise AI applications, Israeli government and public sector AI, Hebrew content generation and analysis. Users working on complex, multi-step problems benefit most from DictaLM 3.0 24B Thinking's ability to maintain coherence across long inputs and produce structured, verifiable outputs.

Like all current AI systems, DictaLM 3.0 24B Thinking has limitations: Hebrew-first focus reduces quality on non-Hebrew tasks; 24B requires 16GB+ VRAM for FP16 self-hosting; Hebrew specialization means a narrower target market than multilingual models. For production deployments, validating outputs against authoritative sources — especially for domain-specific or time-sensitive information — remains best practice. Dicta continues to improve DictaLM 3.0 24B Thinking through ongoing research.

DictaLM 3.0 24B Thinking is available on LITHOVEX AI with a free account. Unlike direct API access — which requires billing setup, API key management, and per-token charges — LITHOVEX AI provides a seamless browser interface with no setup required, making DictaLM 3.0 24B Thinking accessible to students, researchers, developers, and early-stage teams globally.

As of 2026-10-16, DictaLM 3.0 24B Thinking by Dicta demonstrates measurable performance on standard AI evaluation benchmarks. On Hebrew NLU — which measures model performance on specialized evaluation tasks — it achieves 79.3%. These scores place DictaLM 3.0 24B Thinking — at 24B of parameters — in a competitive position within its model class. Note that benchmark scores reflect controlled evaluation conditions; real-world performance varies by task, prompt design, and domain specificity.

Dense transformer, 24B, Hebrew-specialized with thinking mode

The training pipeline for DictaLM 3.0 24B Thinking involves multiple stages: large-scale pretraining on diverse multilingual text and code corpora to develop broad world knowledge, followed by supervised fine-tuning on instruction-response pairs, and alignment techniques such as RLHF or DPO. Each stage refines the model's helpfulness, accuracy, and safety. Dicta's research team — based in Israel — has optimized DictaLM 3.0 24B Thinking for the specific capability profile described above.

Inference efficiency: DictaLM 3.0 24B Thinking at 24B is designed to deliver quality outputs at practical serving speeds. Its context capacity supports complex multi-turn conversations without quality degradation.

Getting started with DictaLM 3.0 24B Thinking on LITHOVEX AI takes under two minutes. Follow these steps:

For best results with DictaLM 3.0 24B Thinking: provide specific, context-rich prompts; include examples when generating code; specify output format when structure matters; break complex tasks into focused sub-questions. Since DictaLM 3.0 24B Thinking supports extended reasoning, prompting it to "think step by step" or "show your reasoning" activates deeper chain-of-thought analysis for math, logic, and complex problems.

Dicta is a leading organization in AI research and large language model development. Their team has built expertise across natural language processing, model architecture design, and alignment research. Dicta's models are used by researchers, developers, and enterprises worldwide. Their commitment to advancing AI capability while maintaining responsible deployment practices has made them a respected contributor to the global AI ecosystem. The organization continues to publish research and release models that push the boundaries of what AI systems can achieve.

Founded: 2021, headquartered in Israel.

Visit the official Dicta website →

DictaLM 3.0 24B Thinking by Dicta competes in a field with several strong alternatives. Below is a detailed side-by-side comparison against commonly listed alternatives, using benchmark data and capability profiles in the catalog.

Both DictaLM 3.0 24B Thinking and the models above are available free on LITHOVEX AI — test them side by side on your real tasks before committing to a subscription with any provider.

DictaLM 3.0 24B Thinking is released under Apache 2.0, which means you can download and run it on your own hardware. Below are the quickest ways to get started.

Hebrew is a significant global language serving Israel's 9 million speakers plus diaspora communities worldwide. DictaLM 3.0 provides Hebrew language AI that genuinely understands the unique morphological and grammatical complexity of Modern Hebrew — critical for Israeli technology companies, government, and academia.

DictaLM 3.0 24B Thinking by Dicta has the following version history and notable updates:

For the complete changelog and latest release notes, visit the official Dicta model page.