UberLLM vs Vercel AI Gateway
Vercel charges each provider's list price and adds team controls; UberLLM routes each request to the cheapest healthy seller of unused compute, so the per-model price can be lower, with no fee per request and a 5% fee only when you add credit. UberLLM also publishes that failed generations cost nothing and lets you cap the price per request.
Right now UberLLM routes 455 models, 191 of them for images, video, music and speech.
UberLLM vs Vercel AI Gateway, side by side
| UberLLM | Vercel AI Gateway | |
|---|---|---|
| How you start | $5 minimum top-up, no subscription | Free monthly credit; paid credits in any amount you choose |
| Free to try | None, prepaid credit | Monthly free credit on a subset of models, with lower rate limits (ends once you buy credits) |
| Fee on top of model prices | None per request; 5% when you add credit | No markup on provider token prices; payment processing fees may apply; some optional add-ons cost extra |
| Works with the OpenAI SDK | Yes | Yes |
| Works with Claude Code and the Anthropic SDK | Yes | Yes |
| Chat and text models | Yes | Yes |
| Image models | Yes | Yes |
| Video models | Yes | Yes |
| Music models | Yes | Yes |
| Same model from many sellers | Yes | Yes |
| Failed generations are free | Yes | Not stated |
| No-code app (no programming) | No | No |
| Run it on your own servers | No | No |
Vercel AI Gateway facts are from its own site as of 28 October 2026 (its pricing page). “Not stated” means Vercel AI Gateway does not publish it. The two companies are independent; Vercel AI Gateway is a trademark of its owner. UberLLM prices are live on the model board.
What Vercel AI Gateway is
A managed gateway from Vercel that gives you one API key for hundreds of models from many providers, with request logs, budgets and provider failover. You can pay with Vercel credits or bring your own provider keys.
How you pay
Prepaid AI Gateway Credits, charged at the provider's list price per request with no markup on tokens, including with your own keys. Some optional add-ons (custom reporting, team-wide provider allowlist, team-wide zero data retention, trace drains) are billed extra, and you cover any payment processing fees.
What UberLLM is
A marketplace for unused AI compute. Each request goes to the cheapest healthy seller for that model, behind one OpenAI- and Anthropic-compatible key.
How you pay
Pay per use at the price shown for each model, with nothing added per request. A 5% fee applies only when you add credit. A generation that fails costs nothing.
Pick UberLLM if
Developers who want the lowest live price for chat, image, video and music models under one key, using the OpenAI or Anthropic SDK they already have.
Pick Vercel AI Gateway if
Teams, especially ones already on Vercel, who want provider list prices, their own provider keys, and built-in budgets, logs and data-retention controls in one managed gateway.
What Vercel AI Gateway does that UberLLM doesn’t
- Bring your own provider keys with no markup
- A monthly free credit on a subset of models
- Team, project and member budgets, and trace export
- Zero data retention routing, provider/model allowlists, embeddings, reranking and realtime voice
Switch in one line
Point the SDK you already use at UberLLM. The same key covers chat, images, video, music and speech.
from openai import OpenAI
client = OpenAI(
base_url="https://api.uberllm.dev/v1", # the only line that changes
api_key="ull_your_key_here",
)
resp = client.chat.completions.create(
model="deepseek/deepseek-v4-flash",
messages=[{"role": "user", "content": "Hello!"}],
)
# the same key makes images (and video, music, speech)
img = client.images.generate(
model="venice/nano-banana-pro",
prompt="a red apple on a white table",
)FAQ
Is UberLLM a Vercel AI Gateway alternative?
Yes, for calling models through one key. Both give you OpenAI-compatible and Anthropic-compatible endpoints for chat, image, video, music and speech models. UberLLM is a marketplace that sends each request to the cheapest healthy seller, while Vercel AI Gateway is a managed gateway that charges provider list prices.
How is UberLLM different from Vercel AI Gateway?
Vercel charges the provider's list price with no token markup and lets you bring your own provider keys, plus budgets and logs for teams. UberLLM sells unused compute from many sellers at the price shown per model, adds no fee per request, charges a 5% fee only when you add credit ($5 minimum), never charges for a failed generation, and lets you cap the price of each request. UberLLM does not support bringing your own provider keys.
Do I need to host my app on Vercel to use either one?
No. Vercel's docs say AI Gateway works from any server, cloud or local machine with an AI Gateway API key. UberLLM also works from anywhere with one API key, including with the OpenAI and Anthropic SDKs and Claude Code.