Improve margins across your portfolio
Up to ~20% reduction* in LLM spend through a single line of code
*Based on average savings across customers using Merge Gateway’s intelligent model routing.
“We had an aggressive timeline for launching our enterprise AI search product. Merge moved quickly in supporting the file storage integrations we needed, which allowed us to meet our target launch date.”
“With Merge Agent Handler, we can bring a broader set of data connectors to our users faster and constantly expand what our customers can do with Perplexity Enterprise Pro.”
“Using Merge, we simply have to check off a box to add a new HRIS integration. This gives us full confidence that we can support any of our customers’ HRIS integration needs over time.”
“Before launching our HRIS integrations, a few team members at Merge came on-site to train our customer-facing teams on marketing, selling and supporting the integrations. This took additional work off of my team’s plate and has helped us take the integrations to market more efficiently.”
“Working with Merge’s Unified API and beautiful React component took less than a sprint to integrate, test, and release.”
“Merge’s post-sales team is truly amazing. They’re fast, responsive, and deeply knowledgeable of HRIS integrations and our use case.”
AI is becoming a top 3 cost, and it’s almost always unoptimized
Every portfolio company independently rebuilds AI infrastructure from scratch. The inefficiency compounds across the portfolio.
Defaults to the most expensive model
Engineering hardcodes one model across all use cases.
AI spend can't be forecasted
Provider invoices don’t break out spend by product, team, or customer. Can’t set limits until after costs spike.
Every portco rebuilds the same infrastructure
Every portco spends months building the same cost controls from scratch. No shared leverage, no compounding value.
One line of code.
Deploy once.
No excuses.
Your portco swaps one URL. Everything else stays the same, including code, SDKs, prompts, product.
Optimization happens automatically
Every request is routed to the lowest-cost model that meets quality requirements.
Savings show up immediately
Reduce LLM spend from the first invoice. No engineering work. No rollout risk.
Standardize across every portco
Same integration. Same playbook. No product changes required.
Turn AI cost optimization into a value creation lever
This is not a one-time optimization.
It’s a playbook you apply across every company.
How a $1M/year AI spend becomes $800K
A mid-market SaaS company spending $1M/year on LLM inference. Here’s what happens after a 1-minute integration.
You find the waste in diligence
Engineering hardcoded every AI call to GPT-4 at $30 per million tokens. Simple tagging and complex reasoning both hit the same model.
Portco swaps one line. Merge does the rest.
Each task gets routed to the cheapest model that maintains quality.
| Task type | % of calls | Routing change | Cost /M tokens |
|---|---|---|---|
| Simple tasks (tagging, extraction) | 50% | GPT-4 → Haiku | $0.80/M |
| Mid-complexity (chat, summaries) | 30% | GPT-4 → GPT-4o-min | $0.30/M |
| Complex reasoning | 20% | GPT-4 → GPT-4 (no change) | $30/M |
Margin expansion from month one
No new headcount. No engineering project. Pure margin improvement from month one.
The more companies you onboard, the more leverage you gain
Merge charges LLM cost plus a small margin. No platform fees.
No per-seat charges. The margin shrinks as you scale.
Portcos | Merge margin | Est. annual savings | |
|---|---|---|---|
Portcos | 1–3 | 5.0% | $320k |
Growth | 4–7 | 4.7% | $893k |
Scale | 8-14 | 4.4% | $1.81M |
Portfolio | 15-24 | 4.2% | $3.24M |
Enterprise | 25+ | 4.0% | $4.2M+ |
See how much margin you can unlock across your portfolio
Start with one portco. See results in the first month. Scale from there.
