D

DhiWise (now Rocket.new)

Surat-born design-to-code platform that rebranded in 2025 to Rocket.new — one of India's first AI-native "vibe-coding" app builders, $15M seed-funded with Palo Alto HQ and $4.5M ARR in its first 3 months

AI App Builder / Code Generation 4.4 / 5 (1 Rating) Free / paid tiers (in flux) Updated May 2026 🇮🇳 India roots, US HQ now

Quick Verdict

If you have been tracking the Indian no-code / low-code space, DhiWise was the design-to-code-conversion company — take a Figma file, generate production-ready Flutter or React code, ship faster. That product is largely a chapter of the past. In 2025 DhiWise rebranded to Rocket (rocket.new), pivoted to a fully AI-first model, launched its new "vibe-coding" platform in beta in June 2025, and raised a $15M seed round in September 2025 led by Salesforce Ventures with Accel and Together Fund. Within three months of launch the new product had crossed 400,000+ users with 10,000+ paid subscribers across 180 countries, hitting $4.5M ARR. The company also moved its headquarters from Surat, Gujarat to Palo Alto, California as part of the pivot, while keeping a significant India engineering presence. For Indian product readers, the new product is best understood as one of India's first credible entries in the same category as Lovable, Bolt, v0 and Replit Agent — describe what you want in natural language, get a working app you can iterate on. If you're searching for DhiWise specifically, this page exists; if you're evaluating "vibe-coding" or AI app builders in 2026, you're really evaluating Rocket.new.

📰 2025 corporate update. DhiWise rebranded to Rocket / Rocket.new in mid-2025 and pivoted entirely to an AI-native app builder. Founders Vishal Virani (CEO), Rahul Shingala and Ameya Sahasrabudhe still lead the company. HQ moved from Surat to Palo Alto, California; India engineering remains in Surat. Total funding through both DhiWise and Rocket eras is approximately $25M. The old DhiWise design-to-code product is being wound down in favour of the new Rocket platform.

AI-driven app generation
4.4
Speed to first working app
4.6
Code-export quality
4.1
India support & team
4.2
Product roadmap stability
3.1

What is DhiWise / Rocket.new?

DhiWise was founded in April 2021 in Surat, Gujarat by Vishal Virani (CEO), Rahul Shingala and Ameya Sahasrabudhe. The original product solved an unglamorous but valuable problem: take a Figma design, parse its component hierarchy, and emit production-grade Flutter or React code on the other side. The "design-to-code" thesis was that the gap between high-fidelity designs and shippable mobile or web apps was being filled by manual engineering work that ought to be automatable. DhiWise built that bridge — and through 2021–2024 grew into one of the more interesting Indian developer-tool stories, raising roughly $10M in earlier rounds from Accel, India Quotient and other investors, and reaching meaningful adoption among Indian and global app development teams.

Then, in 2025, the team made a much bigger bet. The same wave of AI coding tools that produced Lovable, Bolt, v0, Replit Agent and the broader "vibe-coding" category made the old design-to-code thesis look incomplete — users no longer wanted to start from a Figma file; they wanted to start from a sentence. DhiWise launched a new product, Rocket (rocket.new), in beta in June 2025, in which you describe an app in natural language and the platform generates the working app — UI, backend logic, database, and deployment — end-to-end. The new platform progressively replaced the old DhiWise design-to-code surface, and the company formally rebranded.

In September 2025, the rebranded company announced a $15M seed round led by Salesforce Ventures with continuing participation from Accel and Together Fund. Salesforce Ventures invested approximately ₹74.46 Cr ($8.46M) as the lead, Accel added ~₹49.26 Cr ($5.6M), and Together Fund put in ~₹7.67 Cr. Headquarters were officially relocated from Surat to Palo Alto, California, though the India engineering team remained in Surat. By the time of the seed announcement — just three months after launch — Rocket had crossed 400,000 users, 10,000+ paid subscribers in 180 countries, and $4.5M ARR. Total funding across the DhiWise and Rocket eras is approximately $25M.

For Indian product readers, the framing that matters is: this is now squarely an AI-app-generator company competing globally — not an Indian design-to-code tool. The team and origin story are Indian; the customers, positioning and pricing are global. The page you are reading exists primarily for two audiences: (1) people searching for "DhiWise" who may not know about the rebrand, and (2) Indian product teams who want to understand whether to seriously evaluate Rocket vs Lovable / Bolt / v0 / Replit Agent for their next app build.

Capabilities (Rocket.new, 2026)

💬 Natural-language app generation

Describe an app in plain English; Rocket generates a working web app (frontend + backend logic + database schema). Each prompt iteration regenerates or modifies the app rather than starting over. Same category as Lovable / Bolt / v0.

🎨 Figma-to-app continuity

The old DhiWise strength carried forward: import a Figma file as a starting point, then evolve it via natural-language prompts. Useful for design-led teams who already have a Figma library.

🧱 Stack output

Frontend in React (with Tailwind); backend in TypeScript / Node; database in PostgreSQL. Code is downloadable and editable — Rocket is not a black-box vendor lock-in like some no-code platforms.

🚀 Hosted preview & deployment

Each prompt iteration deploys to a Rocket-hosted preview URL instantly; one-click production deployment to Rocket's hosting or export to your own infrastructure (Vercel, Netlify, your own server).

👥 Collaboration & auth

Built-in auth scaffolding (email / OAuth / magic links), team-collaboration on the same Rocket project, and version history of prompt-iterations.

🤖 Agent workflows (in development)

Post-seed roadmap includes agent-style workflows where the Rocket agent can autonomously break a task down, generate multiple iterations, run them, and pick the best. Not fully shipped as of mid-2026.

Pricing & plans (2026)

Pricing is in active flux post-rebrand and we expect it to change over 2026. As of mid-2026 the published tiers on Rocket.new are roughly:

  • Free tier — limited credits per month, public projects only. Useful for evaluation and one-off experiments.
  • Paid (Hobby / Pro) — credit-based pricing in the $20–$50/month range, with credits roughly tied to AI-generation tokens. For Indian buyers that's ~₹1,700–₹4,200/month plus 18% IGST.
  • Team / Business — higher credit limits, team collaboration, custom domains, priority compute. $100–$300/month typical; ~₹8,400–₹25,000/month for Indian buyers.
  • Enterprise — custom contracts for larger teams, on-prem or VPC deployment options, custom SLAs.

The page's previous "₹999/month Business" figure was the old DhiWise pricing and no longer applies. Validate live pricing on rocket.new; credit-based AI pricing is volatile in this category and any number cited in print is likely stale within months.

When DhiWise / Rocket.new is the right call

  1. You want to ship a working prototype this afternoon — vibe-coding tools are unmatched for first-draft speed. Rocket is in the same speed class as Lovable, Bolt and v0.
  2. You are a designer or non-engineer building an MVP — describe what you want, get a working app; iterate via prompts.
  3. You have a Figma design and want to evolve it into a real app — Rocket inherits DhiWise's Figma-import strengths, which most pure-vibe-coding competitors don't have.
  4. You want to support an Indian-founder global product — Rocket's engineering team is still in Surat; founders are Indian.

Rocket is the wrong call when: you need a stable, mature, well-documented platform (the rebrand is recent — the product surface is still settling), you want a polished, dev-friendly traditional codebase you'll commit to long-term (use Cursor + a normal stack instead), or you specifically need the most production-grade AI coding agent at writing-real-code (use Claude Code or Cursor with a real engineer).

Pros & cons

✓ Pros

  • One of India's first credible AI-native app builders ("vibe-coding")
  • Inherits DhiWise's strong Figma-import surface — rare in vibe-coding category
  • Strong recent traction: $4.5M ARR within 3 months of beta launch
  • Backed by Salesforce Ventures + Accel + Together Fund ($15M seed Sept 2025)
  • Indian engineering team in Surat — quality talent at India costs
  • Generated code is exportable; not a black-box lock-in

✗ Cons

  • Recent rebrand + pivot — product still settling; expect churn in feature naming and surface
  • Credit-based pricing model is harder to budget for than seat-based
  • USD billing — INR billing not native
  • Competitive intensity in the vibe-coding category is brutal (Lovable, Bolt, v0, Replit Agent, Cursor, plus a long tail)
  • The old DhiWise design-to-code workflow is being wound down; existing customers may need to migrate

Related insights & playbooks