Verdent
Multi-agent coding platform with 76% SWE-bench Verified score
Overview
Verdent is an agentic coding suite — desktop app (Verdent Deck), VS Code and JetBrains extensions — that splits a coding task into parallel agents, each in an isolated git worktree, with planning and verification modes. Pricing from $19/month, with BYOK and pay-as-you-go options.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Claude, GPT, Gemini, GLM and Kimi model access
- Eco Mode for light usage
- Unified billing
- Priority support
- Same model access as Starter
- Eco Mode for extended workflows
- Same model access as Starter
- Eco Mode for intensive workflows
- Organization-wide controls
- Dedicated onboarding
What you can produce with Verdent
- Verdent Deck desktop app (multi-agent orchestration)
- VS Code and JetBrains extensions
- Plan Mode — converts briefs into requirements and action plans
- Isolated git-worktree workspaces per agent
- BYOK and pay-as-you-go credit options
- Reasoning-level controls (minimal to high)
- Verification mode for output checking
ASEAN Perspective
Verdent in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Verdent's pitch is planning before code: a dedicated Plan Mode turns a vague brief into requirements and an action plan before any agent starts writing, which is a real answer to the one-shot-and-hope failure mode of most coding agents. Verdent Deck orchestrates several parallel agents across isolated git worktrees, and on the company's own SWE-bench Verified run it claims a 76.1% single-attempt resolution rate — a strong number, though it's self-reported rather than independently audited.
Founded by former TikTok algorithms lead Zhijie Chen and former Baidu product lead Xiaochun Liu, the company ships fast: VS Code and JetBrains extensions alongside the desktop app. Pricing runs on a credit system that starts reasonable ($19/month) and gets expensive quickly on Max ($179/month) if you're running agents constantly — credit top-ups add up.
Best for solo developers and small teams who want more structure than a bare coding agent, less overhead than a full engineering-agent platform.
What people say
Verdent was founded by Zhijie Chen, formerly head of algorithms at TikTok, and Xiaochun Liu, formerly head of tech and product at Baidu — a pedigree the company leans on in its marketing, and one borne out in the product's technical ambitions. The core idea is running multiple autonomous agents against different parts of a coding task simultaneously, each in its own git-enabled workspace that merges cleanly back into the main branch, with Plan Mode turning a rough brief into requirements and an action plan before any code gets written.
On the company's own SWE-bench Verified benchmark, Verdent reports a 76.1% single-attempt resolution rate, which it markets as leading among production coding agents — worth treating as a vendor-reported number rather than an independently verified one, since no third-party benchmark run turned up in review coverage.
A hands-on review from a developer using it for real client work (a GitHub Copilot integration proof-of-concept she'd been stuck on for ten hours) singled out Verdent Deck as the standout piece — orchestrating an epic-sized project in under 20 minutes — and described the VS Code extension as producing solid output without needing constant hand-holding. She got real value from the reasoning-level controls (minimal through high, though she mostly stuck to "fast" or "smart") and noted the tool recovered cleanly after she deliberately confused an agent mid-merge, restoring worktree data that could have been lost. It gave her one of the clearer breakthrough moments in that project. Her main criticism was vague status messaging from the agent and no visibility into which underlying LLMs power the system.
Pricing runs on a credit model: a 7-day free trial with 100 credits, then Starter at $19/month (480 credits), Pro at $59/month (1,500 credits), and Max at $179/month (4,500 credits), plus a $20/user/month Teams tier and custom Enterprise pricing. Credit top-ups start around $20 for 340 credits at roughly $0.059/credit and don't expire. Simple tasks run 3-5 credits; complex multi-file refactors can burn 30-50+ credits depending on model choice and whether you run Speed or Quality mode — meaning heavy users can chew through a Starter plan fast and need to budget for Pro or Max.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Verdent's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Verdent unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Verdent directly →
Spotted something out of date? Suggest an update →
More in Code & Dev Tools
