Wire · founder news, decoded · technology
Upstage unveils Solar Open 2, open-weight AI model built for autonomous tasks
◆ Published
23 July 2026
◆ Topic
technology
◆ Sectors
◆ Geography
◆ Source
◆ Verified
Fusion42 · 23 July 2026 · Fusion42 review
Upstage, a South Korean AI company, released Solar Open 2, an open-weight large language model optimised for autonomous agent tasks with 250 billion parameters (15 billion active) and 1 million token context window. The model outperforms DeepSeek V4 Flash, Mistral Medium 3.5, and Cohere Command A+ on agent benchmarks, and beats OpenAI GPT-5.4 mini and Anthropic Claude Haiku 4.5 on Korean language tasks.
This Wire brief sits within Fusion42's coverage of AI Frontier Models and AI Agents. Wire is Fusion42's founder-focused intelligence feed: each story is connected to the funds and startups it names — every one with a live profile on Raise or Scout — so founders can follow the capital and the momentum behind the headline rather than just the headline itself. Wire analysis is one of the live surfaces Arthur reasons over.
◆ The Wire takeaway
An open-weight model built specifically for agents (not chat) now has a proven buyer: Upstage itself, deploying it across Daum, finance, law and health. If you're building agent tools or vertical SaaS that needs a reasoning backbone, you have a Korean-made alternative to Anthropic and OpenAI that runs on two H200s and costs nothing to licence.
◆ Related on Wire
- Upstage Unveils Solar Open2, Targeting Enterprise AI with 250B Model22 July 2026
- Upstage Unveils Solar Open 2: Open-Source AI Agent LLM23 July 2026
- China's Moonshot unveils open-weight AI model challenging US rivals17 July 2026
- Moonshot releases Kimi K3, largest open-weight model to date | AI Weekly18 July 2026
- Chinese company releases world's largest open-source AI model16 July 2026
- China's Moonshot Debuts Kimi K3, a 2.8T-Parameter Open AI Model17 July 2026
◆ Topics
AI Frontier Models · AI Agents · open-weight-models · agent-ai · mixture-of-experts · korean-ai · moe-efficiency