Bensk logo

Bensk

Comparing eight LLMs on 38 Berlin election questions

Share on:

An exploratory, reproducible analysis of current language models responding to the 38 theses for Berlin’s 2026 state election.

Berlin WahLLM – Eight language models and the 2026 Berlin state election | Berlin WahLLM DE EN Berlin WahLLM Who would AI vote for? Eight language models (such as ChatGPT) answer the 38 theses (listed below) from the Wahl-O-Mat for Berlin's 2026 state election. The result at a glance Seven of eight model runs rank the SPD, Greens or Left first within the default party selection. Grok stands out clearly, ranking the FDP first, followed by the CDU and AfD. A snapshot: the analysis is based on one run per model, not stable political positions. 8 models · 38 theses An experiment, not voting advice

Related listings

A free MCP server for facts LLMs get wrong (time zones, holidays, FX)

Live conditions from the weather station on my roof on Long Island, NY.

agents that set themselves up from your GitHub repo

Compare monthly snapshots of Wikipedia articles, inspect changed passages, and check cited sources.

Generate outfit transition videos from a single photo

graph-based MCP memory with an RL-trained forgetting policy