LLM API for Next.js: what 301 makers' products chose
Among 301 makers' products built with Next.js that we know the LLM API of, OpenAI API leads: 222 of them (74%). Next come Claude (123) and Gemini API (40). Next.js makers pick Claude 1.6× and Groq 1.7× as often as makers overall.
Which provider, and whether to go through a gateway. Going direct to a provider gets its newest models and features first and the lowest latency; a gateway gives you one key and one bill for hundreds of models, with fallbacks when a provider is down, for a small markup or a proxy you run. The code that calls the model — an AI SDK or agent framework — is its own decision.
What makers on Next.js chose
| Option | Next.js products | All makers | 2026 launches | Free for | At 10,000 users |
|---|---|---|---|---|---|
| 222 (74%) | 75% | 62% | ~1,000 users | $10/moGPT-6 Luna | |
| 123 (41%) | 26% | 39% | ~1,000 users | $10/moClaude Haiku 5.5 | |
| 40 (13%) | 10% | 14% | ~360 users | $28/moGemini 3.1 Flash-Lite | |
| 20 (7%) | 4% | 4% | ~1,500 users | $6.75/moGPT OSS 20B | |
| 17 (6%) | 7% | 10% | — | — | |
| 6 (2%) | 2% | 0% | — | — | |
| 5 (2%) | 2% | 3% | ~1,700 users | $6.00/moMinistral 3 (3B) | |
| 5 (2%) | 1% | 0% | ~370 users | $27/modeepseek-flash | |
| 3 (1%) | 2% | 3% | — | — | |
| 2 (1%) | 1% | 0% | ~110 users | — | |
| 2 (1%) | 0% | 0% | — | — | |
| MiniMaxModel provider | 1 (0%) | 1% | 0% | ~370 users | — |
| Eden AIGateway (many providers, one API) | 1 (0%) | 0% | 0% | — | — |
| DeepInfraOpen models, hosted | 1 (0%) | 0% | 0% | ~1,600 users | — |
Live makers' products with a source for both: that they're built with Next.js, and what they use for LLM API. A product using two options counts in both; not market share. "All makers" is the option's share among every maker's product that made this decision. "2026 launches" is its share among the 77 Next.js products launched that year: a tool that came first has products from every year since, so its total runs ahead.
Costs are the cheapest plan that fits, from each vendor's pricing page, at the usage our stack planner assumes for a product with that many users. Free for: the users its cheapest plan a business can use stays free for; plans for personal use only, like a non-commercial hobby tier, don't count.
Which to pick, by situation
- You want the strongest general models and the widest ecosystem of examples and integrations
OpenAI API 222 here
Claude 123 here
Gemini API 40 here - Token cost dominates your budget, for example high-volume batch or agent workloads
DeepSeek 5 here
Groq 20 here
Mistral AI 5 here - You want to try or mix many models with one key and one bill, and fall back when a provider is down
OpenRouter 17 here - You want that routing and spend control but self-hosted, with keys and logs on your servers
LiteLLM 3 here
OpenAI API
Model provider222 Next.js products here (74%), against 75% of all makers' products. Free up to about 1,000 users. About $10 a month at 10,000 users (GPT-6 Luna).
llms.txtIn Lovable, Replit, RorkBest forThe broadest model lineup — text, images, speech, embeddings — and the largest ecosystem.
Trade-offFrequent model and API changes to keep up with, and no option to run its hosted models elsewhere.
“Helped us bring intelligence directly into startup operations and workflows. From insights to automation, OpenAI accelerated the AI-native vision behind Nautis.”
- Its models are strong at code generation, reasoning and creative tasks, often chosen after testing alternatives. 4 sources
- The API is reliable and well-priced as the foundation for production features. 4 sources
- It cut usage on the $200 Pro plan and changed limits, frustrating paying subscribers. 4 sources
- Model quality shifts between releases, including a quickly replaced GPT-6 and quietly tweaked reasoning levels. 4 sources
Status page: 5+ incidents in the last 90 days (its own page, 2026-10-19).
“We use OpenAI models for parts of automated scenario generation, evaluation and building some of our pre-defined metrics”
LLM APIMaker says so +1 · source ↗“The most reliable foundation model for fast, high-quality copy generation.”
LLM APIMaker says so +1 · source ↗“Great array of models for different purposes, and love the playground for quick exploration”
LLM APIMaker says so +1 · source ↗“We use OpenAI to create our fine-tuned models.”
LLM APIMaker says so +1 · source ↗“The fastest way to go from “what if?” to “oh wow, it actually works.” Rock-solid APIs, great models, and zero patience required for babysitting infra. We tried alternatives — we shipped with OpenAI.”
LLM APIMaker says so · source ↗“We use GPT-4 internally for writing copy (like this one!) and in our app to give our users the highest quality of GenAI answers possible”
LLM APIMaker says so +1 · source ↗“Lovely documentation for building on the OpenAI Response API.”
LLM APIMaker says so +1 · source ↗“Martin uses the GPT family of foundational AI models to power our agents.”
LLM APIMaker says so · source ↗“Helped us bring intelligence directly into startup operations and workflows. From insights to automation, OpenAI accelerated the AI-native vision behind Nautis.”
LLM APIMaker says so · source ↗“Complex analysis, competitive intelligence, deep research”
LLM APIMaker says so +1 · source ↗
Claude
Model provider123 Next.js products here (41%), against 26% of all makers' products. Free up to about 1,000 users. About $10 a month at 10,000 users (Claude Haiku 5.5).
llms.txtIn Lovable, ReplitBest forCoding, long documents and agent-style tool use.
Trade-offNo image generation or embedding models of its own, so those need a second provider.
“Added depth, reasoning, and reliability to AI-driven experiences on the platform. A key part of helping founders make smarter and faster operational decisions.”
- It is strong at writing code and powers many code generation and coding agent products. 4 sources
- It stays coherent through long, multi-step reasoning and agent workflows. 4 sources
- Outages are frequent, with uptime below three nines across the API and apps. 4 sources
- API pricing is a large multiple of subscription pricing, making heavy agent use expensive. 3 sources
Status page: 36+ incidents in the last 90 days, 17 major (its own page, 2026-10-19).
“We use Claude models for parts of automated scenario generation, evaluation and building some of our pre-defined metrics”
LLM APIMaker says so +1 · source ↗“One of our LLM providers, using multiple for different tasks, along with our own agent skills”
LLM APIMaker says so +1 · source ↗“Modelence App Builder is powered by Opus - world's leading coding model”
LLM APIMaker says so +1 · source ↗“Orchestrator synthesis, email scoring, coaching tips, meeting prep, weekly summaries”
LLM APIMaker says so +1 · source ↗“Claude powers our agents, including the coding agent that opens PRs”
LLM APIMaker says so +1 · source ↗“Reliable SOTA AI models.”
LLM APIMaker says so +1 · source ↗“Claude powers the intelligence behind Giselle's AI workflows. We leverage the full Claude family — Haiku for fast, cost-effective tasks on our free tier, Sonnet for balanced everyday workflows, and Opus for the most demanding reasoning challenges. Extended Thinking lets our workflows reason through multi-step problems, while multimodal capabilities enable understanding PDFs and images. The API's reliability and Claude's unique ability to follow complex instructions make it perfect for orchestrating sophisticated AI workflows. Thank you for building AI that truly thinks!”
LLM APIMaker says so +1 · source ↗“At Nia, we aim to provide multiple LLMs so that it is easy to switch between various large language model providers with one click. Hence, using Anthropic models was a no-brainer for us, as they are very strong in reasoning, and we often share them with our clients”
LLM APIMaker says so · source ↗“AI Apps are powered by Claude 3.7 Sonnet”
LLM APIMaker says so · source ↗“We chose Claude because of Anthropic's pace of development and strong engineering focus. Beyond the major model improvements, it's the attention to the little details that makes building with it smoother and helps us move faster.”
LLM APIMaker says so · source ↗
Gemini API
Model provider40 Next.js products here (13%), against 10% of all makers' products. Free up to about 360 users. About $28 a month at 10,000 users (Gemini 3.1 Flash-Lite).
In Lovable, ReplitBest forVery long context and multimodal input, with a free tier to start.
Trade-offFree-tier prompts may be used to improve Google's products, so paid tier is the one for user data.
“Expanded our multi-model AI capabilities with strong contextual understanding. It helped us explore richer AI workflows across planning, research, and execution.”
- It offers some of the best price-to-performance, with fast, cheap Flash models. 4 sources
- A context window of a million tokens or more handles whole codebases and long documents without extra pipelines. 4 sources
- Getting an API key and paying is confusing, with usage tiers and Google Cloud console hoops. 4 sources
- Models are deprecated abruptly, some without leaving preview or having a replacement. 4 sources
“Google's Gemini models bring unmatched context, multimodal, and image generation capabilities to Giselle. With a 1M token context window, our workflows can process entire codebases, long documents, and even video content. From Gemini 2.5 Flash Lite for fast free-tier tasks, to Gemini 2.5 Flash and Pro for balanced and advanced reasoning, Gemini 3 Pro and Flash for the most demanding agentic workflows, and Gemini 2.5 Flash Image for native image generation — we leverage the full Gemini family. Google Search Grounding keeps responses current, while native multimodal support handles text, images, audio, and video seamlessly. Thank you for pushing the boundaries of AI capability!”
LLM APIMaker says so +1 · source ↗“We chose Gemini 3 for thegro.ai because we needed more than text generation — we needed reasoning. Across real outbound workflows (lead research, prioritization, structured outputs, and human-sounding outreach), Gemini consistently handled long context and multi-step logic better than the alternatives we tested. It gave us strong performance, predictable formatting for Gro IQ, and a cost profile that makes scaling a self-serve AI Sales Co-Pilot viable. For us, it wasn’t about hype — it was about building on a foundation that thinks clearly at scale.”
LLM APIMaker says so · source ↗“The native video understanding is what enabled us to build this brand new UX!”
LLM APIMaker says so · source ↗“Powering next-generation AI experiences with multimodal intelligence and contextual understanding.”
LLM APIMaker says so · source ↗“Gemini is our default model for web agents”
LLM APIMaker says so · source ↗“A top tier AI model thats perfectly balanced for speed, price and quality. Also love the rest of the Google Gemini models!”
LLM APIMaker says so · source ↗“The second brain of AI Phone Screener”
LLM APIMaker says so · source ↗“Expanded our multi-model AI capabilities with strong contextual understanding. It helped us explore richer AI workflows across planning, research, and execution.”
LLM APIMaker says so · source ↗“Gemini's extensive token window makes it our go-to solution for one-shot analysis, enabling efficient and precise insights without the need for iterative processing.”
LLM APIMaker says so · source ↗“Fast tool selection, intent classification, signal categorization”
LLM APIMaker says so · source ↗
Groq
Open models, hosted20 Next.js products here (7%), against 4% of all makers' products. Free up to about 1,500 users. About $6.75 a month at 10,000 users (GPT OSS 20B).
Best forVery low-latency inference on open models, for real-time features.
Trade-offOnly the open models it chooses to host, and no frontier closed models.
“We love using Groq hosting and models. They are a core part of our chat and widget agents. The models are lightning fast and the Groq team is fantastic.”
- Inference is extremely fast, enabling real-time voice and agent workflows. 4 sources
- The free tier is generous, with many models to choose from. 3 sources
“We love using Groq hosting and models. They are a core part of our chat and widget agents. The models are lightning fast and the Groq team is fantastic.”
LLM APIMaker says so +1 · source ↗“Remarkable speed and reliability.”
LLM APIMaker says so · source ↗“We use Groq for ultra-fast inference when analyzing millions of contact records and enriching them with AI. It enables us to run deep research and structured reasoning at speeds that would be impossible on standard GPU setups. This level of performance lets us deliver intelligent outputs in real time, even at scale. We're grateful to the Groq team for building the kind of infrastructure that makes this possible.”
LLM APIMaker says so · source ↗“Groq Chat enhances our LPU's performance significantly, enabling faster inference and improved user interaction.”
LLM APIMaker says so · source ↗“YapIt only exists because of Groq. We needed instant transcription and rapid LLM formatting to make the app feel magical. Groq's whisper model and AI inference speeds are unmatched—they make our voice-to-text pipeline feel like it has zero latency.”
LLM APIMaker says so · source ↗“Super reliable and fast so our users don't have to wait long”
LLM APIMaker says so · source ↗“The fastest Realtime latency is not possible without Groq inference”
LLM APIMaker says so · source ↗“Fastest inference on LLaMa models”
LLM APIMaker says so · source ↗“Super fast and globally stable! Sped up our workflow so much <3”
LLM APIMaker says so · source ↗
OpenRouter
Gateway (many providers, one API)17 Next.js products here (6%), against 7% of all makers' products.
llms.txtIn ReplitBest forTrying many models from many providers with one API key and one bill, without opening an account at each.
Trade-offA fee on top of provider prices, and your traffic and data pass through a third party.
“We needed a simple way to access and switch between different AI models while building Shape's agent. OpenRouter made that much easier during development.”
- One API key and OpenAI-compatible endpoint reach hundreds of models from many providers. 1 source
- Switching and comparing models is easy, which speeds up testing cost and quality. 1 source
- Quality varies by underlying provider, so users pin trusted providers to avoid degraded output. 3 sources
- Caching is less effective because requests can switch providers mid-session. 3 sources
“We use OpenRouter to optimize for maximum uptime for all of the third-party models that power Olive. We built Olive around many models across Anthropic, OpenAI, Google, including even open weight models without sacrificing valuable engineering time.”
LLM APIMaker says so · source ↗“We needed a simple way to access and switch between different AI models while building Shape's agent. OpenRouter made that much easier during development.”
LLM APIMaker says so · source ↗“Used as the LLM gateway powering model access (Gemini, Llama, GPT-4o, Claude).”
LLM APIMaker says so · source ↗“The most flexible API to manage all your AI models”
LLM APIMaker says so · source ↗“We use both. Cloro runs most of our models (GPT, Gemini, Perplexity, Copilot) and is the more economical option there. We added OpenRouter specifically for Grok, since cloro dropped Grok support, and tracking all five engines reliably is the whole point. Different tool per model.”
LLM APIMaker says so · source ↗
Azure OpenAI
Through your cloud6 Next.js products here (2%), against 2% of all makers' products.
MCPBest forOpenAI models inside an Azure account, with regional and EU/US data-zone deployments.
Trade-offMore setup than the OpenAI API (Azure resources, deployments per model), and new models can arrive later.
“OpenAI's API might have temporary failures. Azure's OpenAI service is always the fallback we rely on when the OpenAI's API is unavailable.”
“OpenAI's API might have temporary failures. Azure's OpenAI service is always the fallback we rely on when the OpenAI's API is unavailable.”
LLM APIMaker says so · source ↗“Azure OpenAI is helping to scale our application to all provide AI to all of our users”
LLM APIMaker says so · source ↗“We're using Azure OpenAI to do the document processing, shout out to the Microsoft Startup Hub for their support too.”
LLM APIMaker says so · source ↗“We got accepted into the Microsoft for Startups Founders Hub, which means we now have access to free infrastructure credits—a huge boost for our growth!”
LLM APIMaker says so · source ↗“Thanks to OpenAI for the SSML creation. And Azure Speech for TTS.”
LLM APIMaker says so · source ↗Also chosen, and not seen yet
Fewer Next.js products here: Mistral AI (5), DeepSeek (5), LiteLLM (3), Grok API (2), Vertex AI (2), MiniMax (1), Eden AI (1), DeepInfra (1).
Who switches, and to what
Public GitHub pull requests whose title says "X to Y", by developers in general, not only on Next.js.
- OpenAI API → Gemini API 178+ pull requests · how they moved
- Gemini API → Groq 152 pull requests · how they moved
- OpenAI API → Claude 127 pull requests · how they moved
- Gemini API → Claude 108+ pull requests · how they moved
- Claude → Gemini API 104+ pull requests · how they moved
- Gemini API → OpenAI API 90+ pull requests · how they moved
Getting LLM API right
Put model calls behind one function in your code so switching providers is a one-line change. Log prompts, responses and cost from day one. Set spending limits on every provider account.
- Shipping with no spending limit or per-user rate limit, so one abusive user or a retry loop runs up the bill overnight.
- Hard-coding one model name across the codebase, so moving to a newer or cheaper model means a search-and-replace and a round of surprises.
- Putting the provider API key in client-side code, where anyone can copy it from the browser.
The rest of the Next.js stack
What makers' Next.js products chose for their other decisions, the most common pick for each:
- Hosting: Vercel 1,078 of 1,312
- Transactional Email: Resend 493 of 830
- Database: Supabase 171 of 382
- Payments: Stripe 288 of 359
- Authentication: Clerk 176 of 211
- Web Analytics: Google Analytics 105 of 197
- Product Analytics: PostHog 123 of 168
- Lifecycle Email: Loops 87 of 108
Questions
What do most Next.js apps use for LLM API?
Among 301 makers' products built with Next.js that we know the LLM API of, OpenAI API leads: 222 of them (74%). Next come Claude (123) and Gemini API (40). Next.js makers pick Claude 1.6× and Groq 1.7× as often as makers overall.
What does LLM API cost for a Next.js app with 10,000 users?
On our usage model at 10,000 users, OpenAI API comes to about $10 a month (GPT-6 Luna); Claude comes to about $10 a month (Claude Haiku 5.5); Gemini API comes to about $28 a month (Gemini 3.1 Flash-Lite); Groq comes to about $6.75 a month (GPT OSS 20B). Prices are from each vendor's pricing page.
Which LLM API option stays free longest for a Next.js app?
Mistral AI: on our usage model its cheapest plan stays free up to about 1,700 users.
How this is counted: a product is on Next.js when a live-product source says so (the maker, its website, a customer story or a subprocessor list), and the same for its LLM API. How evidence is collected.