Custom Chutes
Deploys a team's private or fine-tuned model on dedicated GPU capacity behind its own API endpoint, billed per second of GPU time plus a one-time deployment fee.
Find alternatives to Custom Chutes
What it does
- Model hosting
- Model inference
Sources
- Pricing
-
Pricing text read 2026-11-16. Quote: "Private Chutes are billed at the GPU's hourly rate for however long the instance runs, plus a one-time deployment fee equal to 3× the hourly rate at the time of deployment." The enterprise-quote model rests on the Custom Chutes page (https://chutes.ai/custom-chutes), whose Pricing section says enterprise and advanced custom deployments are scoped per workload; that sentence is not on the pricing page.
- Description
-
Company disclosure · 16 Nov 2026
Page text read 2026-11-16 (curl, full text). Page quotes: chutes.ai/custom-chutes: "Custom Chutes: Your Model. Your Compute. Your API."; "Deploy private models for production without running platform infrastructure yourself.".
- What it does
-
Company disclosure · 16 Nov 2026
Page text read 2026-11-16 (curl, full text). Page quotes: chutes.ai/custom-chutes: "Custom Chutes: Your Model. Your Compute. Your API."; "Deploy private models for production without running platform infrastructure yourself.".
- Sold within
-
chutes.ai/pricing (read 2026-11-16) puts Private Chutes on the same pay-as-you-go account as the per-token models but states no access relationship for Custom Chutes, so only the parent name is written.
- Name
-
Company disclosure · 16 Nov 2026
Page text read 2026-11-16 (curl, full text). Page quotes: chutes.ai/custom-chutes: "Custom Chutes: Your Model. Your Compute. Your API."; "Deploy private models for production without running platform infrastructure yourself.".
- URL
-
Company disclosure · 16 Nov 2026
Page text read 2026-11-16 (curl, full text). Page quotes: chutes.ai/custom-chutes: "Custom Chutes: Your Model. Your Compute. Your API."; "Deploy private models for production without running platform infrastructure yourself.".
- Deployment
-
Company disclosure · 16 Nov 2026
The product page names a vendor-run deployment path (no platform infrastructure to run), so saas is the vendor-operated reading. Quote: "Deploy private models for production without running platform infrastructure yourself.".
Also from Chutes
-
Chutes Model API
Chutes serves open-source language, image, video and speech models through an OpenAI-style API and runs customers' own models on confidential GPUs.
Something wrong here? Send a correction — quote this product id: chutes-custom-chutes.