
ModelRunner
ClaimedRun image, video, audio, 3D and text models, paying only per use
Paid plans only Visit WebsiteThe Bottom Line
- Best for
- Developers and product teams adding image, video, audio, 3D or text AI features to their applications.
- Price
- Paid plans only. Compare plans
Key facts
- One API key for image, video, audio, 3D and text models.
- Python and TypeScript SDKs with streaming, webhooks and polling.
- Hosted MCP server with OAuth sign-in for AI assistants.
Not for
Teams that require offline inference or want to operate the model-serving infrastructure themselves.
What is ModelRunner?
Available on: API
Key Features
- Unified API for image, video, audio, 3D, and text models
- Playground for testing models with example prompts
- MCP server for calling models from AI assistants
- Asset library for uploads, inputs, and generated outputs
- Custom model wrappers with selected inputs and prompts
Pricing
ModelRunner offers paid plans. Visit their website for current pricing details.
Reviews

Review ModelRunner, get a free AI guide
Share your experience and we will send you Improve Your Thinking Patterns Using ChatGPT, free.
Best ModelRunner Alternatives
Top alternatives based on features, pricing, and user needs.
Deploy and scale ML models with fast cold starts and dedicated GPUs
Run, fine-tune, and deploy open-source ML models via API
Run generative AI models for image, video, and audio 4x faster with serverless GPUs.
Accelerate your AI with developer-friendly APIs for performance and cost-efficient machine learning inference.
Still deciding?
Most buyers shortlist 2 or 3 tools before committing. Pull a side-by-side comparison or browse the full alternatives shortlist below.
Explore More
ModelRunner FAQ
How does ModelRunner help developers test AI models before integrating them?
How does ModelRunner compare to Replicate for running AI models?
What are the main limitations or trade-offs of using ModelRunner?
Which teams benefit most from using ModelRunner?
How is ModelRunner priced?
Can ModelRunner be used from within an AI assistant workflow?
How do custom endpoints work in ModelRunner?
Source: modelrunner.ai