Tabnine

Enterprise AI dev platform whose Code Review Agent checks PRs against plain-language team rules on GitHub, GitLab, and Bitbucket.

[ Facts ]

Category: AI PR Review
Open source: No
Pricing: Code Assistant platform $39/user/mo, Agentic platform $59/user/mo (annual billing); no free tier listed as of Aug 2026 source
Self-hosted: Yes — full stack — Both plans list SaaS, VPC, on-premises, and air-gapped deployment options.
Platforms: github, gitlab, bitbucket
Model control: Choose from Tabnine's supported models incl. its protected model; private/self-hosted model deployments supported
Last verified: 2026-08-11

Based on public documentation as of 2026-08-11. ✓ documented · ~ partial · ✗ not offered · ? unknown. Methodology on the about page.

Enterprise context engine draws on codebase awareness; no documented multi-repo dependency resolution.

Prebuilt plus fully customizable plain-language rules; changed lines are checked against team rules.

Reviews run in the IDE, via CLI (pre-commit hooks), and on pull requests.

Separate Jira Validation Agent checks code against Jira issue requirements.

No documented learning from review feedback over time.

No sandbox execution or runtime validation documented.

Model switching within a vendor-supported list and on-prem model hosting; no pay-your-own-provider BYOK.

Flags rule deviations with guidance and suggested edits; committable-fix flow not documented.

Analytics dashboard tracks usage and acceptance; no DORA or review-ROI attribution.

Evaluating Tabnine?

Run it through the two-week trial protocol before you commit.