WildBench A
AI benchmark · Reasoning
What is WildBench?
WildBench evaluates LLMs on about 1,024 challenging real-user tasks, scoring responses with checklist-guided LLM judges (WB-Score and WB-Reward) that correlate with human preferences.
General capability mostly
“WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild (v2)”
github.com/allenai/WildBench ↗
Who built WildBench
Led by Allen Institute for AI (AI2).
Tier A
Widely used: a lab reports it, or 50 or more models are scored on it.
How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.
- 61 models scored
- in curated RL directories (2 pts)
On X · unclear, low buzz, from 1 post about WildBench in the last month
Only passing mention in a paper summary on reasoning traces in distilled DeepSeek R1 models; little direct discussion of WildBench itself.
How we got to tier A
- Adoption: 61 models scored. Tier A.
- On X: unclear, low buzz. “Only passing mention in a paper summary on reasoning traces in distilled DeepSeek R1 models; little direct discussion of WildBench itself.” No move.
- In curated RL directories: 2 points, a floor of C.
- Confirmed: a line from its own site or paper places it.
A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.
Where WildBench sits
Its domain on the map, live. Hover a logo for its name; click to open it. Open it on the full map
Next: browse all benchmarks. Entry from the RL Research daily scrape of public sources, 2026-11-14 snapshot.