Does Tabnine Train on Your Data? Plan-by-Plan Answer with Quotes
Tabnine states it does not train its models on customer code, but its private fine-tuned enterprise models are described as pretrained on private code.
Last checked Nov 9, 20265 verified quotes3 source documentsRead by Claude Haiku 5.5 · quotes verified by code
Trains on your data
DPA
Data location
Retention
Plan by plan
“All plans” means the vendor’s documents do not distinguish plans for this question.
| Question | Enterprise |
|---|---|
| Trains on your data | Trains on your data |
| How to opt out | Not stated |
| Data retention | Code discarded immediately after response; metrics and logs kept one week |
| Where data is stored | US (AWS North Virginia, GCP Iowa, Heroku Virginia) |
| DPA available | Not stated |
| Subprocessors | AWS, Google Cloud, Heroku (hosting); no third-party AI model APIs |
| Certifications claimed | Not stated |
| Human review of your data | Not stated |
Trains on your data
- EnterpriseTrains on your data
How to opt out
- EnterpriseNot stated
Data retention
- EnterpriseCode discarded immediately after response; metrics and logs kept one week
Where data is stored
- EnterpriseUS (AWS North Virginia, GCP Iowa, Heroku Virginia)
DPA available
- EnterpriseNot stated
Subprocessors
- EnterpriseAWS, Google Cloud, Heroku (hosting); no third-party AI model APIs
Certifications claimed
- EnterpriseNot stated
Human review of your data
- EnterpriseNot stated
What Tabnine says, word for word
Open a field to read the vendor’s sentence, its source and our confidence.
Trains on your dataAll plansDoes not train
“Tabnine does not train its models on your code”
Tabnine's code completion and chat models are trained only on permissive open source code.
Trains on your dataEnterpriseTrains on your data
“Private fine-tuned AI models are pretrained on private code by Tabnine and are only accessible by your team members and stored on your private setup.”
Customer-specific private fine-tuned models; scope of private code used is unclear.
How to opt outAll plansNot stated
The vendor’s documents we read do not say this. We show “Not stated” instead of guessing.
Data retentionAll plansCode discarded immediately after response; metrics and logs kept one week
“The moment that output is generated, the code is discarded and is never stored.”
Personal information kept no longer than necessary, per privacy policy.
Where data is storedAll plansUS (AWS North Virginia, GCP Iowa, Heroku Virginia)
“The Tabnine SaaS platform utilizes Amazon Web Services (AWS) and Google Cloud Platform (GCP) to host our cloud-based capabilities.”
SaaS hosting; main datacenters only, no EU residency stated.
DPA availableAll plansNot stated
The vendor’s documents we read do not say this. We show “Not stated” instead of guessing.
SubprocessorsAll plansAWS, Google Cloud, Heroku (hosting); no third-party AI model APIs
“No third-party APIs are used.”
Google Cloud Identity also handles user identity for cloud solutions.
Certifications claimedAll plansNot stated
The model’s supporting quote was not found word for word in the source, so this field is shown as Not stated.
Human review of your dataAll plansNot stated
The vendor’s documents we read do not say this. We show “Not stated” instead of guessing.
Change history
No changes detected since we started tracking Tabnine on Nov 9, 2026 (version 1). We re-check its documents regularly.
Sources we read
- Privacy policyPrivacy Policy | Tabnine: The AI code assistant that you control
- Privacy policyPrivacy | Tabnine Docs
- Trust / securitySecurity | Tabnine Docs
More on Tabnine
Other ai coding assistants
AI Coding Assistants that don’t train on your data →Informational, not legal advice. Every verdict is what the vendor’s own documents state, quoted word for word. Read the cited source before you decide, and check your own contract: negotiated terms can differ.