Chutes.ai: Run open-source models without managing GPUs
Chutes.ai is a serverless compute platform for deploying, scaling and running open-source AI models, including LLMs, image, video, speech and music models.
What is Chutes.ai?
Chutes.ai provides serverless compute for open-source AI models. Developers call a chat completions endpoint or browse the model catalog to run LLMs, embedding models, content moderation models, and image, video, speech and music models. The platform says it adds new state-of-the-art open-source models shortly after release, and all featured models run on confidential TEE compute. Beyond public inference, teams can deploy their own private chute from the CLI, choosing a GPU class through a NodeSelector and paying by the second while the instance runs. Billing is pay-as-you-go per million tokens with no subscription required, or a monthly Plus or Pro plan that bundles a daily request quota with a discount on per-token rates beyond it.
Focus areas: Model API, Model Inference.
Who is Chutes.ai for?
- AI developers building inference-backed products
- Teams deploying private or fine-tuned models on dedicated GPUs
- Product teams that need image, video, speech or music models alongside LLMs
Is Chutes.ai worth it?
A pay-as-you-go inference and GPU platform for teams that want open-source models without operating their own infrastructure
Where Chutes.ai is strong
- Pay-as-you-go token pricing with no subscription required
- Private chute deployments on confidential TEE GPU capacity, billed per second
- Catalog spans LLMs plus image, video, speech and music models
Where it falls short
- Free trial availability is not stated in the source material
- Plan quota sizes are not specified; the pricing page only links to plan limits
How do you get started with Chutes.ai?
- 01Create an account and get an API key
- 02Send a request to the chat completions endpoint with your chosen model
- 03For a dedicated workload, build and deploy a private chute with the CLI
Chutes.ai: frequently asked questions
How does Chutes.ai billing work?
Public inference is billed per million tokens with no subscription, minimum or markup. Private chutes are billed by the second at the GPU's hourly rate, plus a one-time deployment fee equal to three times the hourly rate.
What does a Private Chute cost?
The self-serve TEE GPU class listed is an RTX Pro 6000 at $1.80 per hour, with a one-time deployment fee of $5.40. You pay only while the instance runs, and idle instances shut down automatically.
Do I need a subscription to use Chutes.ai?
No. Pay-as-you-go is available, and the optional Plus ($10/mo) and Pro ($20/mo) plans bundle a daily request quota with a discount on per-token rates beyond that quota.
Which models can I run on Chutes.ai?
The catalog includes LLMs such as Kimi, GLM, DeepSeek, Qwen, Gemma, MiniMax and Mistral, plus embedding and content moderation models and image, video, speech and music models. You can also bring your own code and run any open model.
Chutes.ai alternatives
The closest tools in our directory, matched on category, focus areas and audience.
| Tool | Best for | Free plan | Starting price | Focus areas |
|---|---|---|---|---|
| Chutes.ai | Developers running open-source model inference at scale | No | — | Model API, Model Inference |
| PeriFlow | Teams serving open-weight models to production agents | No | — | Model API, Model Inference |
| Novita | Developers running model inference, GPU jobs and agent workloads | No | — | Model API |
| bytez | Developers deploying open source models via an API | Unknown | — | Model Inference |
| iCreat AI | ICreat AI - Your All-in-One AI Image & Video Creator, Create More, Spend Less. | Unknown | — | Model API |
| Synexa | Synexa AI: Deploy AI models with one line of code, cost-effective, scalable GPU infrastructure. | Unknown | — | Model API |
| Samaira AI | Samaira AI is an AI inference platform offering scalable and secure deployment of machine learning models, simplifying the process for… | Unknown | — | Model Inference |
| Modal | Developers running Python and GPU inference workloads in the cloud | Unknown | — | Model Inference |
| Beam | Beam is a tool that enables developers to rapidly develop AI projects without managing infrastructure. | Unknown | — | Model Inference |
| Vidgo AI | Vidgo AI generates videos from images and text using AI-powered tools. | Unknown | — | Image API, Model API, Video API |
| ModelsLab | Developers who want many generative models behind one API | No | $21/mo | — |

