Want a 51% off Vexp coupon code for vexp.dev? Then click to unlock the discount promo offers below before they expire:
Latest Vexp Coupon Codes & Promo Discount Deals
- Save 51% off with this limited-time Vexp discount code:
1. What Is Vexp?
Vexp is a tool that helps AI coding agents work with your code. Their website calls it “the reliability layer.” It gives the agent the right context before it writes code. It then runs a check before the agent says it’s done.
The problem it targets is easy to relate to. Agents scan files, guess at dependencies, and burn tokens rediscovering your codebase every session.
Vexp fixes that in three steps:
- Index. It maps your code into a dependency graph.
- Traverse. It ranks the code that matters for your task.
- Capsule. It hands the agent a small, focused bundle.
Everything runs on your machine. Their website says there are zero network calls.
2. Who Created Vexp?
Their website credits Omniselect di Nicola Alessi. That name appears in the footer copyright, next to an Italian VAT number. It points to an Italian business run by Nicola Alessi.
I couldn’t find a longer founder story on the page I reviewed. Their website doesn’t list a team or an About section.
It does show a solid public footprint:
- 33,000+ installs across the VS Code Marketplace, npm, and Open VSX
- Public docs, changelog, and blog
- GitHub and LinkedIn links
- Legal pages, including Privacy, Terms, and a Refund Policy
For a tool that reads your code, I’d like more about the people behind it. The visible paper trail is a good start.
3. How Much Does Vexp Normally Cost?
Vexp uses flat pricing. Their website says there are no credits and no surprises.
| Plan | Price | Limits | Highlights |
|---|---|---|---|
| Starter | $0 forever | 2,000 nodes, 1 repo, 20 calls/day | Context search, session memory |
| Pro (most popular) | $19/month | 50,000 nodes, 3 repos, 1,000 calls/day | Impact analysis, CodeLens, email support |
| Team | $29/user/month | Unlimited nodes and repos, 1,500 calls/day per seat | Shared workspace, analytics, priority support |
| Enterprise SDK | Custom | Capacity licensing | REST API, CI gate, audit log, SLA |
An annual option exists too. I’d start free, then move to Pro if you outgrow the 2,000-node cap.
4. What Are The Benefits Of Using Vexp?
Here are the biggest wins I see:
- Fewer tokens. Their example shows 8,247 tokens shrinking to 2,140. That is 74% fewer.
- Lower cost per turn. Your agent skips file-by-file exploration.
- Better answers. The agent sees precise structure, not a flood of unused code.
- A built-in check. The agent’s work gets verified before it says done.
- Privacy. Your code never leaves your machine.
- Fast. Full index in under 15 seconds. Queries land in under 500 ms (P95).
It works across 17 agents and 34 languages. That breadth matters to me, because most teams mix tools.
5. What Are The Main Features Of Vexp?
Their website says to reach for run_pipeline first. It runs context search, impact analysis, and memory recall in one call. Then come 11 specialized tools:
- verify_done: a mechanical completion check, free on every plan
- get_context_capsule: context bounded to your token budget
- get_impact_graph: every caller and dependent that breaks if you change a symbol
- search_logic_flow: trace the path between two functions
- get_skeleton: file structure without bodies
- Memory tools: get_session_context, search_memory, and save_observation
- Setup tools: workspace_setup, index_status, and expand_vexp_ref
It uses tree-sitter to parse code and stores everything in local SQLite. Secrets are filtered out by default.
6. What Companies And Brands Use Vexp?
Their website doesn’t name any company customers. I won’t invent any.
Here is what I can confirm from their website:
- 33,000+ installs across VS Code Marketplace, npm, and Open VSX
- 17 agents supported, including Claude Code, Codex, and Opencode
- Editors: VS Code, Cursor, and Windsurf
- 34 languages covered
It shows one short quote from an unnamed subscriber. That person says they like the approach to searching large codebases.
The Enterprise SDK is aimed at platform teams. They run agent fleets or CI pipelines. I’d ask them for named references before a team rollout.
7. How Have They Benefited From It?
Their website shares test results instead of user stories. These are the numbers:
- 81.7% of tasks resolved on SWE-bench Pro (597 of 731)
- $1.52 cost per task, using Claude Opus 5 tokens
- 74% fewer tokens in their context example
They compare it to Anthropic’s reported figures. Claude Fable 5 is listed at 80.3%. Claude Opus 5 alone is listed at 79.2%.
The run used the official Scale AI harness, and they say every session is published. It was a single run with a 95% range of 78.7 to 84.3. These are their own results, so I’d test it on your own code.
8. Does Vexp Come With A Satisfaction Guarantee?
I didn’t see a money-back guarantee on their website. The footer does link a Refund Policy page, but I couldn’t read its terms. Check it before you pay.
They cut your risk in other ways:
- Free Starter plan. It is free forever, with no account and no credit card.
- Zero network calls. Your code stays on your machine.
- Secret filtering is on by default.
- verify_done checks your agent’s work, free on every plan.
- Promo codes are accepted on Pro and Team.
My tip: test the free plan on one small project first.
9. What Other Deals Can You Kindly Bestow Upon Me?
Check out our daily deals page to browse all the latest deals and discounts on marketing, design, and creative tools and tech. Most of these deals are like a fart in the wind – only available for a limited time. So if you want to save money on something you were going to buy anyway, bookmark the page and check it out daily…