Gupshup vs MSG91: Which is Better in 2026?

The short answer

For most Indian teams that live on SMS OTPs, wallet top-ups, and “just ship the notification,” MSG91 is the sensible default. Gupshup pulls ahead when you are already treating WhatsApp as infrastructure (banks, large EdTech stacks, anything that needs ACE-style agents and multichannel glue without apologising for the UI).

Where Gupshup actually wins

BFSI and other regulated shops do not buy messaging because the dashboard looks pretty; they buy because the vendor has seen a crore messages fail for stupid reasons and has runbooks. Gupshup has been in that game long enough that “enterprise-grade” is not marketing copy here — it is the expectation that your SLAs and audit trail bits exist. ACE (their agent stack) matters if you are wiring LLM-style flows into WhatsApp at volume, not if you are running three broadcast templates a week.

• High-volume multichannel without five vendors: WhatsApp + RCS + SMS + voice + email from one CPaaS contract, which matters when procurement says “single PO” and your architects say “single observability story.”
• Deep APIs for weird integrations: Salesforce/HubSpot/Zoho/LeadSquared exist, but the real win is custom piping into core banking or LMS rails where “just use Zapier” died in 2019.
• Campaign + catalog / commerce flows when you are building conversational journeys that are closer to product than marketing (OTP is table stakes; the journey is not).

Counter-example: A 4-person D2C brand on Shopify that mainly wants broadcasts and quick replies will feel over-sold and under-delivered — too much sales process, not enough instant gratification.

Where MSG91 actually wins

Indore is not a flex; it is a signal that someone picked up the phone in IST and understood why DLT template drama is not “optional homework.” MSG91’s natural home is Indian product teams that already think in per-SMS economics and per-OTP reliability, then bolt WhatsApp on without treating the whole company as an enterprise RFP.

• Wallet prepaid + transparent per-message math for finance teams that freak out when a usage bill lands like a monsoon.

That is it for the bullet list on purpose — because the real advantage is economic and operational: cheapest serious entry for SMS at scale, OTP with Android auto-read, RCS in the same breath, and “we already use Razorpay for something else” level integrations. If your bottleneck is developers who want APIs and your marketing team can live with a spartan console, you are home.

Counter-example: You need polished marketer-first WhatsApp campaign tooling and hate top-up workflows — MSG91 will feel like a backend, not a growth suite.

Pricing, in INR, no spin

Neither vendor publishes a public menu that lets you model next quarter’s burn without talking to someone or clicking through prepaid calculators.

Illustrative maths (simplified, because both sit on Meta’s per-conversation rules for WhatsApp):

Say you run a logistics marketplace doing ₹50 lakh GMV/month with an average ticket of ₹1,200 — that is roughly 4,167 paid orders (rounding is cruel; reality is messier). Assume 60% need a transactional WhatsApp update (dispatch, delivery, one follow-up) and 20% also get a marketing nudge in the same month — about 2,500 + 833 ≈ 3,333 “touch” conversations if you count conservatively (in practice, retries and opt-outs move this).

• Meta conversation charges (India rates change; check current pricing sheets) might land in a band of ₹0.25–₹1.50 per business-initiated conversation depending on category and whether Meta’s free windows apply — call it ₹1,000–₹5,000/day at that volume if you are sloppy, half if your templates are tight. That stack is not Gupshup vs MSG91; it is Meta’s line item on both.
• Platform fee: MSG91 states roughly ₹0.10/msg platform on WhatsApp atop Meta; Gupshup is often quoted around ₹0.10–₹1/msg in self-serve tiers depending on volume and packaging. On 3.3 lakh platform-touched messages (if you foolishly message per event instead of batching), ₹0.10 alone is ₹33,000; at ₹0.50 it is ₹1,65,000/month — which is why seniors yell about payload design.

SMS side (where MSG91 usually wins arithmetic): Transactional SMS in the ₹0.15–₹0.30 range per the sheet you pasted — if OTP + alerts run 8 lakh SMS/month, that is ₹1,20,000–₹2,40,000 before GST, plus DLT scrub headaches that cost time (which is money).

Hidden / annoying costs to model:

• GST on invoices (Gupshup: India invoicing with GST; MSG91: GST on wallet recharges and usage) — your finance person cares more than engineering does, until filing day.
• Wallet float on MSG91: cash sits idle; working-capital nerds notice.
• Sales-led minimums on Gupshup for enterprise packaging, possible setup or professional services (no universal public number — assume ₹0–₹few lakhs until they send the SOW).
• FX / settlement if any sub-vendor bills [USD] (always re-check contract schedules — some global pass-throughs still sting on Forex + TDS paperwork).
• Add-on modules (campaign analytics, advanced routing) that unlock only after you are committed.
• Time cost: Gupshup’s longer setup versus MSG91’s “top up and push” — that is ₹X lakh in engineering hours if you are honest about blended cost.

Net: On OTP-heavy + prepaid culture builds, MSG91 usually wins the spreadsheet until you hit enterprise procurement requirements. On multichannel + AI agent + SLA builds, Gupshup’s invoice may hurt, but replacing three vendors sometimes still nets positive.

What we’d actually use each for

If you are a 12-person D2C team on Shopify with ~₹40L MRR and your main pain is abandoned carts + delivery updates + the occasional festival blast — MSG91 is the boring pick (wallet, SMS backup, WhatsApp on Meta rails). Gupshup only if your CTO already standardized on it for something else.

If you are a BFSI stack where a failed OTP is front-page news (touch wood), RBI tokenisation drama is lunch-table talk, and you need WhatsApp + SMS + voice with audit-friendly operations — Gupshup is the one banks put on shortlists; MSG91 can still win technically, but you will fight perception battles in committee.

If you are Indian SaaS with 40 lakh MAU and your cost line item is “SMS OTP + WhatsApp transactional,” MSG91 tends to stay until you need ACE-grade conversational agents — then you evaluate Gupshup like you’d evaluate any serious platform migration: with fear.

Indian fit (GST, UPI, IST, support)

Both bill in INR for practical Indian use and play nice with UPI in the “Indian buyer” sense (wallet top-ups, settlements, local payment rails — verify at signup because product surfaces change).

GST: Expect GST on Indian invoices from Gupshup; MSG91 will GST your wallet recharges and usage. Neither is a magic “skip compliance” button; e-invoicing thresholds still apply to your business when you cross them (the tool does not absolve your CA).

Support hours: Gupshup lists Mon–Sat, 9am–7pm IST; MSG91 Mon–Sat, 10am–7pm IST. Not 24×7 on paper for either in that snippet — plan for off-hours incidents like adults.

Honest bit: Gupshup’s SF / Mumbai story sometimes shows up as “global first, India second” in meetings (fair or unfair). MSG91 reads as India-native in tone — which matters when you need someone to prepone a template approval call because your DLT header is misfiled.

Migration: what’ll bite you

Gupshup → MSG91: You will redo webhook endpoints and template namespaces (Meta-side migrations are never copy-paste). Campaign history exports may be incomplete for analytics parity; expect to rebuild dashboards. Enterprise contract lock-ins or minimum commits can make the finance team twitch. Integrations like LeadSquared-specific glue might need rewriting compared to Zoho/SFDC paths you already have.

MSG91 → Gupshup: Wallet psychology flips to contract/billing rhythms; your finance team must stop thinking purely prepaid. Bot flows built around MSG91’s modules may not map 1:1 to ACE patterns — developers whine, rightly. DLT/SMS header portability is its own circus; do not assume “export CSV and pray.”

Both directions: customer opt-in records and per-channel consent are legal assets — treat migrations like a banking cutover, not a weekend hack.

What we’d pick

We would run MSG91 for the median Indian product that bleeds money on OTP volume and wants CPaaS sanity without a quarter-long procurement opera. We would pick Gupshup when WhatsApp is mission-critical infrastructure with AI agents, multichannel mandates, and procurement that wants one throat to choke (and will pay for it).

If your founder is arguing at 11pm, ask them this: are we optimising for ₹per message or ₹per incident — because those are two different religions.

Things people actually ask

“Bhai is MSG91 really cheaper if we do ₹2 cr/yr on SMS + WA combined?”

Usually yes on SMS at scale because their ₹0.15–0.30 band and wallet discipline map cleanly to Indian volume deals. On WhatsApp, both eat Meta conversation pricing — your savings come from platform fee tiers and engineering efficiency, not magic.

“Do I need to redo my GST invoice template if we switch vendors?”

You will issue/receive invoices under new GSTIN lines and line items; your e-invoicing/E-way obligations depend on your turnover thresholds, not the CPaaS logo. Ask your CA (seriously).

“Will UPI receipt mismatches break wallet top-ups?”

Sometimes. UPI payments can sit in pending states; finance needs a reconciliation SOP. Not vendor-unique, but MSG91’s wallet model surfaces the pain earlier.

“Is Gupshup ‘foreign’ support hell?”

Not exactly — they list IST hours and have Mumbai presence, but enterprise timelines can feel sales-heavy compared to “top up and ping support” vibes.

“We only care about broadcasts. Who wins?”

Neither is as cosy as AiSensy/Wati-style marketer tools (your JSON even admits it). Between these two, MSG91 tends to be less painful if your team is developer-led; Gupshup if you need multichannel + scale and will live in APIs.

“MSG91 docs look mid. Is that going to slow us?”

Probably a few days of dev time versus a slick SaaS. If you are competent with REST + webhooks, you’ll survive; if you need hand-holding UI, you’ll complain in Slack.

“Can we pass RBI-ish scrutiny either way?”

Vendors give you tools and logs; you still own data flows, consent, and vendor due diligence. Banks pick Gupshup often for history; MSG91 still runs serious workloads — your risk team’s checklist matters more than my opinion.

“What breaks first in migration — templates or webhooks?”

Templates (Meta approval + category friction) and webhook idempotency together — double events will duplicate customer messages unless you design for it.

“Do we need ACE / conversational AI at all?”

If your “AI” is three if-else branches, no. If you are routing support tickets and sale qualification on WhatsApp at lakhs of conversations, start a pilot before the board slide decks get emotional.