Go EuropeAIdirectory

Evidence-backed European AI register

Rasa

Reviewed identity, roles, model listings, pricing and evidence for Rasa.

Dataset compiled 2026-08-28 · evidence over inference

HQ Europe: Nomodel labinference provider2 model listings0 token-price rows

Reviewed identity

Headquarters scope
headquarters outside Europe
Headquarters country
US
Headquarters city
San Francisco
Legal entity
GmbH (Berlin) with a US Inc. entity
Ownership
VC-backed private
Founded
2016
Confidence
medium
Last verified
2026-08-22

Links

Official website

Headquarters evidence

Open interactive profile

Headquarters evidence

Headquarters San Francisco, USA, with engineering in Berlin, Germany

Official AI info page gives San Francisco as headquarters and Berlin only as engineering location.

Evidence boundaries

Organisation listings identify a reviewed canonical record. Headquarters is independent of incorporation, registered offices, ownership, hosting and runtime. Model listings show an association in the source register; they do not by themselves establish model creation, current API availability, hosting location or infrastructure ownership.

Model listings

  1. Rasa DIET / TED classifiers and dialogue policies (pre-CALM NLU stack)catalogued with this organisation · origin not asserted
    Type
    text
    Parameters
    small (task-specific transformers)
    Context
    unknown
    Languages
    multilingual
    Licence
    Apache-2.0 (Rasa Open Source)
    Release
    2019–2023
  2. Rasa CALM (Conversational AI with Language Models) — command generator, incl. fine-tuned small modelscatalogued with this organisation · origin not asserted
    Type
    text
    Parameters
    down to 8B (fine-tuned Llama-class for low-latency voice)
    Context
    unknown
    Languages
    multilingual
    Licence
    Rasa Pro commercial licence (CALM); fine-tuning recipes documented
    Release
    2024

Profile note

Berlin-founded and the most widely deployed open-source conversational AI framework in the world before the LLM era; its DIET/TED models were genuinely its own. With CALM it now uses an LLM for understanding while keeping business logic in code, and supports self-hosted fine-tuned small models (down to Llama 8B) for voice latency. One of the few in this segment that publishes an actual entry price for the paid tier (~$35k/year).

Record sources

  1. https://rasa.com/pricing
  2. https://rasa.com/calm
  3. https://rasa.com/docs/learn/concepts/calm/
  4. https://github.com/RasaHQ/rasa
  5. https://rasa.com/