
Crun
One API for every top AI model, video through to LLM
About Crun
Crun consolidates an integration problem that has quietly become substantial. A product needing video generation, image generation, speech and a language model currently means four vendor relationships, four billing arrangements, four sets of credentials and four APIs that behave differently. Worse, the best model in each category changes every few months, so any of those integrations may need replacing. Crun puts them behind one API and one credit balance, which turns swapping models into a parameter change rather than an engineering project.
Coverage is genuinely broad rather than a handful of headline models. Video includes Seedance 2.5 and 2.0, Veo 3.1, MiniMax H3, Wan 2.7, HappyHorse, Kling 3.0, Gemini Omni, Vidu Q3, Grok Imagine and Runway Gen-4. Image includes Nano Banana 2, Seedream 5.0 Pro, GPT Image 2, Flux 2 and Kontext, Imagen 4, Qwen Image 3.0 and Z-Image Turbo. Audio covers Qwen3 TTS and a Suno API. Language models span GPT 5.6 and 5.5, Gemini 3.1 Pro, Claude Fable 5, Sonnet 5, Opus 4.8, Sonnet 4.6 and Haiku 4.5, GLM 5.2 and Doubao Seed 2.1. Pricing is prepaid credits with rates improving substantially at volume.
Top up a prepaid balance and integrate once against the Crun API rather than against each provider individually. Choose the model per request, so a video generation, an image, a speech synthesis and a chat completion all run through the same integration with the same credentials. Because the model is a parameter rather than an integration, adopting a newly released model or moving off one that has become expensive requires changing a string. Credit rates improve with the size of the top-up, running from a starting rate through progressively better rates at higher volumes, and higher-tier plans unlock lower per-model costs on top of that.
- •400+ Models Behind One API - Video, image, audio and language models from a single integration
- •Model as a Parameter - Switching models is a string change rather than an integration project
- •Current Frontier Coverage - Includes recent releases across every category rather than a stale roster
- •Four Modalities - Video, image, audio and text from one credential and one balance
- •Prepaid Credit Balance - Spend bounded by what you deposit, with no surprise invoices
- •Volume Discounts - Credit rates improve with top-up size, and higher tiers reduce per-model costs
- •Unified Billing - One relationship instead of separate accounts with every model provider
- •Low Entry Point - Start from a small top-up rather than a minimum commitment
For developers and product teams building multimodal AI features who would otherwise maintain several provider integrations in parallel. It suits anyone whose product spans generation types, such as an app that produces both images and video, where the alternative is two or three vendor relationships. Teams that want to benchmark models against each other benefit from being able to do so without new integrations each time. Very high volume users on a single model may still get better rates direct from that provider, so the gateway is worth comparing rather than assuming.
Pricing
Usage based, no flat monthly plan
Prepaid credits from $5 (200 credits per dollar), improving to 206 at $50, 212 at $500 and 224 at $1,250 - higher-tier plans also unlock lower per-model costs
From the vendor pricing page, 2026-09-22














