The most intelligent LLM on a single GPU
upstage
company
Verified
AI & ML interests
Building intelligence for the future of work.
Recent Activity
Papers
View all Papers Organization Card
At Upstage, our mission is to build intelligence for the future of work. By building advanced language models and document processing engines, we help organizations redefine the way they work β making processes smarter, more efficient, and ultimately more human-centered.
- Start using our models: https://console.upstage.ai/
- Join the team: https://upstage.ai/careers
spaces 4
pinned
Runtime error
Agents
Featured
570
Open Ko-LLM Leaderboard
π
Explore and filter language model benchmark results
Sleeping
Agents
43
Ko-FreshQA Leaderboard
π
Explore, submit, and download Korean QA leaderboard data
Running
Agents
13
Evalverse
π
Generate LLM evaluation reports and visual analyses
models 25
upstage/Solar-Open2-250B
Text Generation β’ 250B β’ Updated β’ 2.78k β’ 570
upstage/TinySolar-187m-4k
0.2B β’ Updated β’ 29 β’ 6
upstage/TinySolar-111m-4k
0.1B β’ Updated β’ 12 β’ 4
upstage/Solar-Open-100B
Text Generation β’ 103B β’ Updated β’ 7.73k β’ 490
upstage/solar-pro2-tokenizer
Updated β’ 2
upstage/solar-pro3-tokenizer
Updated β’ 2
upstage/TFLOP
Updated β’ 30 β’ 2
upstage/solar-pro-tokenizer
Updated β’ 4
upstage/solar-1-mini-tokenizer
Updated β’ 14
upstage/solar-pro-preview-instruct
Text Generation β’ 22B β’ Updated β’ 37.9k β’ 462
datasets 6
upstage/Pretraining_Dataset
Viewer β’ Updated β’ 60k β’ 167 β’ 3
upstage/dp-bench
Updated β’ 1.42k β’ 85
upstage/wmlu-koarenahard-0-1
Viewer β’ Updated β’ 500 β’ 14 β’ 1
upstage/wmlu-arenahard-2-0
Viewer β’ Updated β’ 500 β’ 10 β’ 1
upstage/CReSt
Viewer β’ Updated β’ 2.25k β’ 54 β’ 7
upstage/TFLOP-dataset
Updated β’ 59 β’ 3