Stockmark-13b
Stockmark Inc. · Japan
- Params
- 13B
- Context
- 4,096
- Min VRAM
- 8 GB
- GGUF
- No
Excellent Japanese fluency, strong instruction following.
Region slug: japanAI models, vendors, and hardware buying realities from Japan. Curated by myaihardware.com.
Japan is a critical node for AI builders because it is one of the few places where high-bandwidth memory (HBM) and advanced packaging substrates are still manufactured, making it a bottleneck for GPU supply chains. Its dominant model families, like the open-source LLM series from Stability AI Japan and local fine-tuned variants of Llama and Qwen, reflect a strong preference for efficiency and domain-specific adaptation over raw scale. An underrated angle is that Japan's aging industrial and robotic infrastructure creates a massive early-adopter demand for on-device AI that must run on legacy hardware, forcing builders to optimize for limited memory and thermal constraints before the rest of the world catches up.
METI has allocated significant funding to accelerate domestic large language model development as part of Japan's sovereign AI push, aiming to reduce reliance on foreign models and secure strategic autonomy. The government's AI Strategy 2024-2025 prioritizes compute infrastructure and evaluation benchmarks while confronting the intrinsic challenge of Japanese tokenization, where complex script mixing and morphological ambiguity inflate token counts and degrade efficiency. This funding directly targets companies and research consortia building models with native script-aware tokenizers to overcome the linguistic bottlenecks that handcuff generic architectures.
Filter by vendor, license, or parameter range.
Stockmark Inc. · Japan
Excellent Japanese fluency, strong instruction following.
Region slug: japanStockmark Inc. · Japan
Massive knowledge capacity, state-of-the-art Japanese generation.
Region slug: japanKarakuri AI · Japan
Fluent conversational Japanese, good adherence to prompts.
Region slug: japanKarakuri AI · Japan
Long context window, efficient inference comparable to 13B dense model.
Region slug: japanTokyo Institute of Technology / NII · Japan
Lightweight, strong Japanese base model.
Region slug: japanTokyo Institute of Technology / NII · Japan
Balanced performance, good for fine-tuning.
Region slug: japanTokyo Institute of Technology / NII · Japan
Highest Japanese capability among Swallow versions, strong zero-shot.
Region slug: japanTokyo Institute of Technology / NII · Japan
Very fast inference, permissive license, good Japanese.
Region slug: japanTokyo Institute of Technology / NII · Japan
Long context, efficient MoE, strong Japanese performance.
Region slug: japanTokyo Institute of Technology / NII · Japan
Superior Japanese understanding with latest Llama 3 architecture.
Region slug: japanELYZA Inc. · Japan
Very high Japanese chat quality, widely used in Japan.
Region slug: japanELYZA Inc. · Japan
Fast, lightweight, excellent Japanese for its size.
Region slug: japanPreferred Networks · Japan
Trained from scratch on high-quality Japanese data, solid base model.
Region slug: japanPreferred Networks · Japan
Exceptional knowledge and reasoning, near GPT-4 on Japanese.
Region slug: japanSakana AI · Japan
High Japanese performance via evolutionary model merging, cost-effective 7B model.
Region slug: japanSakana AI · Japan
Japanese multimodal capabilities, text and image understanding.
Region slug: japanSB Intuitions · Japan
Excellent Japanese language model, cost-effective 7B.
Region slug: japanSB Intuitions · Japan
Larger size improves reasoning and Japanese fluency.
Region slug: japanSB Intuitions · Japan
Top-tier Japanese LLM performance.
Region slug: japanFujitsu · Japan
Trained on Fugaku supercomputer, strong Japanese capabilities.
Region slug: japanRakuten · Japan
Commercial-friendly Apache 2.0, strong Japanese performance.
Region slug: japanRakuten · Japan
Good at conversational Japanese tasks.
Region slug: japanNTT · Japan
Efficient, good performance in Japanese, designed for enterprise use.
Region slug: japanCyberAgent · Japan
Open source, strong Japanese generation.
Region slug: japanCyberAgent · Japan
Largest open Japanese chat model from CyberAgent, good instruction following.
Region slug: japanCyberAgent · Japan
Pioneering open Japanese LLM, good base.
Region slug: japanLLM-jp (NII, etc.) · Japan
Open source, strong Japanese benchmarks, community-driven.
Region slug: japanLLM-jp (NII, etc.) · Japan
Largest open Japanese LLM, strong performance.
Region slug: japanrinna · Japan
Based on Qwen2.5, strong Japanese due to continuous pretraining and fine-tuning.
Region slug: japanllama.cpp, MLPerf, and inference leaderboards.
Will this model fit on your GPU?
Local LLM setup, homelab, mini-PC builds.
How we benchmark, verify ASINs, and source data.
New to local AI? Start here.
Sourcing & commerce disclosure
Model metadata on this page is compiled from each vendor's public HuggingFace repo, model card, or research paper at time of indexing. Performance figures (KMMLU, MMLU, HAERAE, AraBench, IndicGEM, etc.) link directly to source benchmarks where available; missing or partial values are shown as ", ".
Outbound hardware links (GPUs, mini-PCs, accelerators) are Amazon Associate links using store ID fredoline-20. As an Amazon Associate we earn from qualifying purchases. Prices and availability shown elsewhere on the site are snapshots, not live quotes — verify on Amazon before purchasing. See About / disclosures for the full policy.