What Tabnine Is Actually For
Tabnine is an AI coding assistant, priced at $39 to $59 per user per month with no free tier, that was acquired by testing company Tricentis on July 30, 2026 for its Enterprise Context Engine. Existing support continues, but neither company has announced a product roadmap for the coding-assistant line itself.
The short version
Tricentis, a software testing company, acquired Tabnine on July 30, 2026, for its Enterprise Context Engine, not for its coding assistant business. Pricing and support haven't changed, but neither company has said what happens to Tabnine's $39-$59 coding-assistant plans next.
Tricentis, a software testing company, announced on July 30, 2026 that it had acquired Tabnine. Tabnine's pricing page still lists two coding-assistant plans in September 2026, Code Assistant at $39 a seat a month and Agentic Platform at $59, both billed annually, with nothing cheaper on the page.
Every review site still calls Tabnine "the privacy-first coding assistant," and none of the ones ranking for that search have updated to say Tabnine is no longer an independent company. Tricentis did not buy the coding assistant business. It bought the Enterprise Context Engine underneath it, the knowledge graph that maps a codebase, and is folding that into its own Agentic Quality Engineering Platform for software testing, not for writing code.
That leaves the product every reviewer is still describing in a strange position: still for sale, still priced at $39 to $59 a seat, and owned by a company whose actual business is testing software rather than helping developers write it.

Tabnine's own pricing page still carries the acquisition banner above its unchanged $39 and $59 plans, captured 6 September 2026.
What actually changed
Chris du Toit, writing for Tabnine's leadership on the company's own blog, said "existing customers will continue to receive support for the products they use today." A Tricentis vice president described the appeal as giving AI testing agents shared memory and context so they stop shipping untested code, according to a report by DevOps.com.
Neither company disclosed financial terms. Calcalistech reported the deal was estimated at tens of millions of dollars, an unofficial figure neither side has confirmed. Neither announcement says what happens to Code Assistant or Agentic Platform as standalone products once the Context Engine is absorbed into a testing platform.
Tabnine's pricing page itself has not moved. It still lists the same two seat prices, support for models from five major labs, and the same on-premises and air-gapped deployment options it listed before the deal. What it does not list, and what neither company's announcement addresses, is a product roadmap.
Who this still serves, for now
Today, nothing has changed for the buyer Tabnine was already built for: healthcare, finance, defense and other regulated engineering teams that cannot let code leave their own infrastructure. Support continues, the $39 and $59 tiers still work the same way, and Tabnine's Code Review Agent, named Best Innovation in AI Coding at the 2025 AI TechAwards for catching pull-request defects and hardcoded secrets, still ships as part of Agentic Platform.
What that reader does not get, as of this piece, is any statement about whether Tabnine will keep shipping new coding-assistant features under an owner whose core product is test automation.
Who it doesn't serve
It never made sense for an individual developer or a five-person startup to pay $39 for Tabnine over Cursor's $20 Pro plan or GitHub Copilot's $10 one, both of which include a free tier Tabnine dropped in April 2025. The acquisition adds a second reason to look elsewhere: a coding assistant with an unclear product future is a worse bet for a team with no compliance reason to need it specifically.
That gap is exactly where cheaper, narrower tools compete instead. Code Genius starts free and tops out at $49 a seat. Trylle runs sandboxed coding agents that ship pull requests for a few dollars a month in usage credits rather than a per-seat license.
Sennu AI reviews nothing but Salesforce code and sells no self-serve plan at all, proof that pull-request review is fragmenting into narrow tools rather than consolidating under one platform. None of the three needed Tabnine to exit. They were already priced for the market Tabnine left in 2025, and now they compete with whatever Tricentis decides to keep funding, not with a coding-assistant company chasing the same customers they are.
The turn
The obvious objection: acquisitions like this happen constantly, and most don't kill the acquired product. Existing Tabnine customers have a specific written commitment that support continues, which is more than most acquisitions offer on day one. That's a fair point, and it's why nothing in this piece is a prediction that Tabnine shuts down.
What makes this deal different from a coding-tool company buying another coding-tool company is the buyer's actual business. Tricentis sells software testing, not developer tooling, and it said plainly what it wanted: the context graph, for testing agents. Continued support today is not the same commitment as continued investment in new features tomorrow, for a product line outside the acquirer's core market.
Nobody asks whether Cursor's coding assistant has a future when Cursor raises a funding round, because coding assistants are the entire business. That question is legitimate here precisely because it wasn't.
What to watch
Watch for a public product roadmap. If Tricentis or Tabnine names concrete plans for Code Assistant or Agentic Platform, new pricing, new features, or a stated release cadence, before the end of Q1 2027, that is a sign the coding-assistant line survives as a real product. If the year passes with only "continued support" and no roadmap, that is the quieter way a product gets wound down without anyone announcing it.
The question worth asking about Tabnine was never really whether it is private enough. Every reviewer already agrees it is. It is whether a coding assistant can keep being built once the company that owns it is in a different business, and a team unsure which side of that line it falls on can run its own requirements through Smart Match rather than guess from a pricing page that has not caught up with its own acquisition.
Frequently asked questions
Who bought Tabnine?
Tricentis, an enterprise software testing company, announced on July 30, 2026 that it had acquired Tabnine. Tricentis wanted Tabnine's Enterprise Context Engine, a knowledge graph of a codebase, to power its own AI testing agents, not to expand into coding assistants.
Does Tabnine still work the same way after the Tricentis acquisition?
Yes, as of September 2026. Tabnine's pricing page still lists the same two plans, Code Assistant at $39 a seat a month and Agentic Platform at $59, and Tabnine's own announcement says existing customers keep support. Neither company has stated a product roadmap beyond that.
Does Tabnine have a free tier?
No. Tabnine retired its free Basic plan along with its $9 Dev and $12 Pro tiers in April 2025, according to a 2026 review by TechBotHQ, and has not reintroduced one. Its cheapest plan as of September 2026 is $39 a seat a month.
Is Tabnine cheaper than GitHub Copilot or Cursor?
No. GitHub Copilot has a free tier and a $10 Pro plan; Cursor has a free Hobby tier and a $20 Pro plan. Tabnine's cheapest seat, at $39 a month, costs roughly double Cursor's paid tier and close to four times Copilot's, with no free option at all.
Who is Tabnine actually built for?
Regulated and privacy-bound engineering teams, in industries like healthcare, finance and defense, that need to keep code off third-party servers. Its Code Review Agent won Best Innovation in AI Coding at the 2025 AI TechAwards, and its deployment options include on-premises and air-gapped installs.
Covered in this guide
- Tabnine: Enterprise AI coding assistant with privacy-first architecture, agentic workflows, and flexible deployment
- Cursor: Cursor is an AI code editor built on VS Code, widely adopted across large enterprise engineering teams, with Agent Mode, Tab completion, and Cloud Agents.
- Code Genius: AI code generator saves 35% of development time with test generation, refactoring, and documentation for React, Vue, and Tailwind CSS.
- Sennu AI: AI tool that reviews Salesforce pull requests using org-specific context and automates QA testing, claiming 99% reliability; no published pricing.
- Trylle: Trylle is a Git platform that runs AI coding agents in a sandbox and ships results as pull requests, with paid plans offering up to 100 GB of storage.
Sources
Still deciding?
This guide covers a handful of options. Smart Match checks every listing in the directory against how you actually work and what you can spend, then hands you the shortlist and the reason behind each pick.
Start Smart MatchRelated guides
- AI Development Services in 2026: Which Layer You Actually NeedBuyer's guideHow to pick, across a category
- The AI Tool Ecosystem in 2026: Buy the Meter, Not the CategoryBuyer's guideHow to pick, across a category
- Best AI Chatbots in 2026: Pick by the Job, Not the LeaderboardBuyer's guideHow to pick, across a category
- Best AI Coding Assistants in 2026: Pick the Job, Not the BrandBuyer's guideHow to pick, across a category
- Claude Max Used to Die by Wednesday. Anthropic's Fix Expires August 19.AnalysisWhat changed and who it affects
- Cursor Composer 2.5 vs. Claude Code: Which AI Coding Agent Should You Use in 2026?ComparisonHead-to-head, with a verdict