Solar LLM
Upstage AI's efficient 10.7B model with depth upscaling — outperforms Mistral and Llama at its size.
Overview
Solar is a 10.7 billion parameter large language model developed by Upstage AI, a South Korean AI company, using a novel technique called Depth Upscaling. Rather than training a large model from scratch, Depth Upscaling extends the architecture of smaller pre-trained models by adding and combining transformer layers, then continues training the combined model. This approach produces a model with better performance than both constituent models.
Solar 10.7B achieved top performance on the HuggingFace Open LLM Leaderboard in December 2023, outperforming Mistral 7B and Llama 2 13B despite being only 10.7B parameters. The efficient architecture means it can be deployed on a single consumer GPU while delivering quality competitive with larger models.
Upstage AI is notable as a South Korean company making significant contributions to open-source LLM development, with particular strength in Korean language models. Solar supports both English and Korean effectively, making it valuable for ASEAN applications requiring Korean language capability alongside English.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 8 May 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
ASEAN Perspective
Solar LLM in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Solar (Upstage) is a family of efficient open large language models from Korea, with the Apache-2.0 Solar 10.7B famously topping the Hugging Face Open LLM Leaderboard at release thanks to its 'depth up-scaling' technique that delivers strong performance at modest size. Upstage also offers managed Solar APIs and document-AI products, giving both open weights and a commercial path.
It suits teams wanting a compact, commercially-usable open model for self-hosting, fine-tuning or cost-efficient inference, and it carries real APAC relevance given Upstage's Korean base and Asian-language focus. Caveats: the headline open model has been overtaken by newer open releases, so check current Solar versions for state-of-the-art needs; self-hosting requires MLOps capability. The managed API and SEA-region accessibility (Upstage operates across APAC) are pluses, and documentation for the API is reasonable.
Notable facts
- Solar 10.7B was trained by a South Korean company and outperformed all European and American models of similar size, demonstrating APAC's growing contributions to frontier AI.
- The Depth Upscaling technique used for Solar allows models to be built more efficiently by combining and extending existing pre-trained architectures.
- Solar was the first non-Western model to achieve #1 on the HuggingFace Open LLM Leaderboard.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including Solar LLM's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Solar LLM unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Solar LLM directly →
Spotted something out of date? Suggest an update →
Alternatives to Solar LLM
More in LLMs & Chat
