Kolibri 1

Aleph Alpha llm

Aleph Alpha German/English MoE reasoning model trained from scratch on 20T tokens.

overall ⌄

43.0

75 of 125 ranked

price · $/M tokens

self-hosted

open weights, no first-party API

to run it yourself ⌄

47 GB

minimum · 187 GB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmarkpublishedmeasured
MMLU-Pro80 reasoning—
GPQA84.3 reasoning—
AIME 202596.9 reasoning—
SWE-bench66.4 —
HLE21.5 reasoning—
BrowseComp29.4 —

Lineage

No recorded lineage — root or standalone model.

view in full graph →

Metadata

orgAleph Alpha
released2026-10-03
params78B-A3.5B
licenseapache-2.0
hf idAleph-Alpha/Kolibri-1