cogvlm2-llama3-chat-19b
963±15
948CI979
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,318
- Last sync
- 11 min11 minutes ago
All sources syncing normally
- Arena · Text· 11 minutes ago
- Arena · WebDev· 11 minutes ago
- Arena · Vision· 11 minutes ago
- Arena · Agent· 11 minutes ago
- GPQA Diamond· 49 minutes ago
- SWE-bench· 49 minutes ago
- Terminal-Bench· 49 minutes ago
- ARC-AGI-2· 49 minutes ago
Filter by org:
Released:
One grid, every column. Pick axes in the Axes panel — one axis ranks like a single board, several blend into a weighted composite, all of them read as the index. Rank is assigned before any filter and never renumbers.
| Model | Org | ± CI | Swing | Add to compare | |||||
|---|---|---|---|---|---|---|---|---|---|
| 146 | cogvlm2-llama3-chat-19b | zai | May 2024 | — | 963 | 948–979 | — |
Showing Vision Arena exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.