Every AI model, one key, dirty cheap
Sign in with Google or a one-time code sent by email.
Made on MartLLM
Veo 3 Fast$0.12 / sec
One key for every major model
OpenAI
Anthropic
Google
Kling
ByteDance
Runway
DeepSeek
xAI
Alibaba
Black Forest Labs
Moonshot AI
Qwen
Meta
Mistral
MiniMax
Zhipu AI
PixVerse
Luma
ElevenLabs
NVIDIA
- 99.9%
- Max off list, live
- $0
- Monthly fee
- $0
- On failed requests
Where the capacity comes from
MartLLM does not run models. Each request goes to a seller that serves the model through its own account with an API provider (for Claude, GPT and Gemini: the model maker, or a provider such as Amazon Bedrock, Snowflake, Venice AI or OpenRouter), at a price the seller sets below list. Sellers must own or be authorized to resell the access they list under their provider’s terms; stolen, leaked, shared, free-trial and evaluation keys are prohibited. A request that fails is not billed. How it works and why prices can be far below list
One key. Every kind of model.
base_url = https://api.martllm.com/v1
DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-1017$0.0072 / $0.0319$0.15 / $0.67−95%
Claude Haiku 5.5anthropic/claude-haiku-5.5$0.018 / $0.09$0.10 / $0.50−82%
GPT-6.1 Solopenai/gpt-6.1-sol$0.0608 / $0.30$2.00 / $10.00−96%
Gemini 3.8 Flashgoogle/gemini-3.8-flash$0.0748 / $0.37$0.75 / $3.75−90%
per 1M tokens, in / out
Coding
Run Claude, GPT, Gemini and DeepSeek for less
Chat, code and agents through the OpenAI and Anthropic SDKs you already use. You change the base URL and nothing else. Each request goes to the cheapest seller that is working, and a failed request costs nothing.
Video
Make video with Veo, Kling, Seedance, Runway and Wan
Text-to-video and image-to-video from the leading video models, billed per second at the price of the cheapest seller that is working. A clip that fails costs nothing.




Image
Generate and edit images
Nano Banana, FLUX, Seedream and more, through the same image endpoint the OpenAI SDK already calls. These pictures were made with Nano Banana 2.1.
Ads
Start from ads that already sell
Real ads that kept running for weeks or reached millions of views. Advertisers keep paying for an ad only while it makes money. Pick a type, then make your own version with the video and image models here.
The law of the market
More sellers, lower prices
Sellers compete for every request, and the cheapest one that is working gets it. That is why prices here sit below list price, and why they keep falling as more sellers join.
Transparent pricing, built for scale
Leading AI models at a fraction of the cost. One balance, no hidden fees, and you never pay for a failed generation.
Top saving, live
up to 99.9%
off the list price, routed automatically
Failed tasks
$0
zero-risk billing — a failure costs nothing
Fee on requests
None
No MartLLM fee on requests. Adding credit by card costs 5.5% + $0.50.
Models you can call now
Developer-first integration
Up and running in minutes on one unified, OpenAI- and Anthropic-compatible API.
Keep your SDK
Point the base URL at MartLLM and use the OpenAI or Anthropic SDK you already have. Nothing else in your code changes.
base_url =
https://api.martllm.com/v1
api_key = ull_•••••
One id, every backend
Switch models by changing the model id — Claude, GPT, Gemini, DeepSeek, or an image, video or music model. Every request routes to the cheapest healthy seller.
Stream, poll or webhook
Text streams back token by token. Image and video jobs return a task you poll, or point a webhook at your app to be notified the moment the result is ready.
Why developers choose MartLLM
We handle the complexity of many providers and sellers, so you can focus entirely on building.
Lower-cost access
Pay less than list price
Sellers set their prices below list price, and cheaper offers that are working get more of the requests. Many models land well under list price; the deepest savings show live in the table above.
Unified API
One key for every model
Reach Claude, GPT, Gemini, DeepSeek and more — plus image, video and music models — through one OpenAI- and Anthropic-compatible endpoint. Change the model id, not your backend.
Reliability
Routing that reroutes
When a seller is slow or down, the request moves to the next healthy one automatically. Async task tracking and webhook callbacks keep high-volume media workflows moving.
Zero-risk billing
Never pay for a failure
A request or a generation that fails costs nothing. One balance covers everything, with no monthly fee and no hidden charges.
Drop-in
Integrate once
Consistent request and task patterns across text, image, audio and video, with clear docs — so your team can adopt a new model without rebuilding anything.
Playground
Try before you build
Compare model outputs and tune parameters in the dashboard before you write a line of production code — the same API, in a console.
Sell your capacity
List capacity you may resell
Hold provider credits whose plan permits resale, at a provider on our approved list? List them on the market and earn when requests route to you. Another provider can be added on request, after we test it.
Point your coding agent at MartLLM
Two environment variables point Claude Code at MartLLM. The same key reaches every model. Nothing else in your setup changes.
Keep the exact models Claude Code ships with. Opus stays Opus, Sonnet stays Sonnet — routed through the market instead of billed at list.
Set it up by hand
# Make a key at https://martllm.com/dashboard/keys, then:
export ANTHROPIC_BASE_URL=https://api.martllm.com
export ANTHROPIC_AUTH_TOKEN=ull_your_key
claudeLasts for this terminal. To keep it, put the same values in the env block of ~/.claude/settings.json. Guide
Prefer an installer? The Claude Code page has the script, its SHA-256 and how to check it before you run it.
Models and current prices
Every model you can buy right now, at its current price. A request costs what its provider charged, with no MartLLM fee.
| LLM | $0.00286 /M in $0.00999 /M out | −99.8% $1.40 /M in $4.40 /M out Official | |
| LLM | $0.00753 /M in $0.0282 /M out | −96.2% $0.20 /M in $0.75 /M out Official | |
| LLM | $0.00718 /M in $0.0319 /M out | −95.2% $0.15 /M in $0.666 /M out Official | |
| LLM | $0.1798 /M in $0.899 /M out | −91% $2.00 /M in $10.00 /M out Official | |
| LLM | $0.0749 /M in $0.3743 /M out | −90% $0.75 /M in $3.75 /M out Official | |
| LLM | $0.678 /M in $3.39 /M out | −83.1% $4.00 /M in $20.00 /M out Official | |
| LLM | $0.1996 /M in $0.998 /M out | −90% $2.00 /M in $10.00 /M out Official | |
| LLM | $0.000146 /M in $0.000456 /M out | −99.9% $0.15 /M in $0.47 /M out Official | |
| LLM | $0.00168 /M in $0.00415 /M out | −98.6% $0.12 /M in $0.475 /M out Official | |
| LLM | $0.0634 /M in $0.1268 /M out | −95.1% $1.30 /M in $2.60 /M out Official |
Frequently asked questions
Everything you need to know about integrating and billing with MartLLM.
What is MartLLM?
MartLLM is one OpenAI- and Anthropic-compatible API key for Claude, GPT, Gemini, DeepSeek and other text, image, video and music models, usually below list price. Each request is served by a third-party provider and billed what that provider charged.
How is it cheaper than going direct?
Sellers set their own prices as a discount off the model’s list price, and cheaper offers that are working get more of the requests. Sellers must have the right to resell what they list under their provider’s terms. The live price per model is in the pricing table; how it works is at martllm.com/how-it-works.
Which models can I use?
Leading LLMs (Claude, GPT, Gemini, DeepSeek and others) alongside image, video and music models. The market table lists every model that is live right now with its current price.
Do I pay for failed requests?
No. A request or a generation that fails costs $0. You are only billed for results you actually receive.
Is there a fee, a monthly fee or a minimum?
No MartLLM fee on requests. Adding credit by card costs 5.5% + $0.50. There is no monthly fee and no minimum spend: add credit from $5, keep one balance and pay as you go. Credit does not expire while your account is open.
Is it OpenAI-compatible?
Yes. Set the base URL to our endpoint and keep your existing OpenAI SDK; the Anthropic-compatible endpoint works the same way. You switch models by changing the model id.
What happens if a seller goes offline?
Your request is routed to the next healthy seller for that model automatically. For image and video jobs, task tracking and webhook callbacks mean an interruption on one seller doesn’t lose your job.
How do you handle my prompts and data?
Keys are scoped and revocable, and requests are encrypted in transit. Your traffic is used to serve your requests — not to train models.
How do I get started?
Create a free account, add credit (from $5), generate a key, point your SDK’s base URL at MartLLM, and send your first request. New accounts start at $0, so add credit before your first call.
Is MartLLM the same as UberLLM?
Yes. MartLLM was called UberLLM until November 2026. Same service, same accounts and API keys; uberllm.dev and api.uberllm.dev keep working.
Does MartLLM check that I get the model I asked for?
Yes, for sellers who list on MartLLM. They can list only through a provider on our approved list, and the provider’s model must be the model listed. Every offer is tested and reviewed by our team before any buyer reaches it, and keeps being tested on live requests. A seller caught serving a different model is banned, and the buyers it served are refunded from its unpaid earnings. What the checks catch and what they can’t: martllm.com/verification.
Can I sell my own capacity?
Yes, if you have the right to resell it under your provider’s terms and that provider is on our approved list (OpenRouter, Groq, Together AI, Fireworks AI, DeepSeek, Amazon Bedrock and others). Another provider can be added on request, after we test it. Your own server can’t be listed. See martllm.com/sell.
Start building with the world’s top AI models today
One key for every model, priced below list — and you never pay for a request that fails.
