RankingIA v0.6.1
EN

Mistral 7B Instruct v0.2open

Mistral

No score: no objective benchmark
Input / output
—
US$ per 1M tokens
Context
—
tokens
Release
—
release date
Weights
Open
 

Scores by source

Scores by sourcepercentile = % of ranked models it beats
MetricValuePercentileSource
Arena text (Elo) weight 2 1090 1% LMArena

Compare with other models

At the sources