API compatibility
Inceptron documents an OpenAI-compatible chat completions API.
Scope: Inference API
Source: Inceptron official models and pricingInceptron
5 models in the provider catalogue
About
Inceptron is a Swedish inference API provider with a documented model catalogue, EU/EEA customer-content processing, and an OpenAI-compatible API.
Direct answer
Inceptron states that it does not use Customer Content to train its foundation or fine-tuned models by default, and that zero retention is enabled by default for prompts and outputs unless logging is enabled or otherwise agreed.
Data & compliance
What the provider documents
Data residency
Inceptron states that customer content processed through its services remains exclusively within the EU or EEA.
Scope: Customer Content
Source: Inceptron official privacy and data policyHeadquarters
Inceptron AB identifies its legal address in Lund, Sweden.
Scope: legal_entity
Source: Inceptron official terms of serviceInfrastructure ownership
Inceptron states that it runs on third-party cloud infrastructure while remaining responsible for its service obligations.
Scope: Service infrastructure
Source: Inceptron official terms of servicePrompt retention
Inceptron states that zero retention is enabled by default: prompts and outputs are not logged or stored unless the customer enables logging or agrees otherwise.
Scope: API prompts and outputs
Source: Inceptron official terms of serviceTraining use
Inceptron states that Customer Content is not used to train foundation or fine-tuned models unless the customer opts in or signs a separate agreement.
Scope: API Customer Content
Source: Inceptron official terms of serviceInfrastructure
Regions, facilities, and capacity
Headquarters, compute locations, and data centres are separate facts. A missing public count is shown as not disclosed rather than inferred as zero.
Regional availability
- European Uniondata residency
Infrastructure metrics
Data-centre countNot independently disclosed
No verified public facility count was found in the reviewed sources.
Physical locations
Headquarterscity
Lund, Skåne, SE
Legal-entity address published in the provider terms; this is not presented as a data-centre location.
Source: Inceptron official terms of serviceCatalogue
Available models (5)
DeepSeek-V4 Flash 0731
modelInput
$0.13/MTok
Output
$0.28/MTok
Provider catalogue entry; model-specific regional evidence is not mapped.
Pricing sourceGLM-5.2
modelInput
$1.20/MTok
Output
$4.20/MTok
Provider catalogue entry; model-specific regional evidence is not mapped.
Pricing sourceKimi-K2.6
262k contextInput
$0.73/MTok
Output
$3.50/MTok
Provider catalogue entry; model-specific regional evidence is not mapped.
Pricing sourceKimi-K2.7 Code
262k contextInput
$0.75/MTok
Output
$3.50/MTok
Provider catalogue entry; model-specific regional evidence is not mapped.
Pricing sourceMiniMax-M2.5
196k contextInput
$0.15/MTok
Output
$0.90/MTok
Provider catalogue entry; model-specific regional evidence is not mapped.
Pricing sourceProvider details
What Inceptron provides
Inceptron offers compiler-accelerated inference through an OpenAI-compatible API and a curated model library.
What is documented about infrastructure
The provider publishes EU/EEA processing language, but does not publish a verified public data-centre count in the reviewed sources.
Answer engine
Common questions
Does Inceptron train on user prompts?
Its terms say Customer Content is not used to train foundation or fine-tuned models by default; explicit opt-in or a separate agreement is required.
Does Inceptron retain prompts and outputs?
Its terms say zero retention is enabled by default and prompts and outputs are not logged or stored unless logging is enabled or otherwise agreed.
Where is Inceptron headquartered?
Inceptron AB lists a legal address in Lund, Sweden.
Evidence
Reviewed sources
Facts and prices can change. Verify the linked provider source before making a procurement or residency decision.