Netlify Brings Open models in the AI Gateway to Production Workflows.

Netlify is adding hundreds of open models to its AI Gateway, giving developers a lower-cost fallback when premium AI usage starts consuming too much of a team’s monthly credit allowance. The company says the expanded catalog includes models from DeepSeek, Qwen and Kimi, among others.

The update targets a familiar problem for teams running AI-powered production workflows. Frontier models can deliver strong results, but their higher usage costs can quickly drain a credit-based budget. In a LinkedIn announcement, Netlify argued that capable open models cost a fraction of frontier pricing, allowing automated work to continue when credits are running low.

That does not mean every workload should immediately move to the cheapest model. It does give teams another lever to control costs without shutting down features or waiting for the next billing period.

AI Gateway provides a managed layer between a Netlify application and model providers. Expanding that layer to open models means developers can access a broader range of models without building a separate provider integration for every option. Netlify’s official announcement presents the catalog as suitable for production use, rather than limiting it to experiments or local development.

The practical options now include:

• Routing routine tasks to lower-cost models, while reserving frontier models for complex prompts
• Keeping background automations operating as a monthly credit allowance gets tight
• Testing the same prompt across more models to compare output quality, latency and cost
• Reducing dependence on a single model provider within an application’s architecture

Netlify has also broadened access through OpenRouter, as previously covered by Create With. The combination gives builders a larger catalog behind the gateway, although the announcement does not provide a model-by-model breakdown of pricing or performance.

For a small business, the most useful pattern may be selective substitution. A customer-facing assistant or complex coding task could continue using a frontier model, while classification, summarisation, extraction and recurring internal jobs move to less expensive open models. The right split will depend on the required response quality and how predictable the workload is.

The central benefit is continuity. If a team has several automated jobs in the queue near the end of a billing period, switching suitable jobs to open models can make the remaining credits last longer. Netlify is not describing the feature as free or unlimited, only as substantially less expensive than frontier alternatives.

Teams will still need to evaluate model behaviour before changing production routing. Open models can vary considerably in instruction following, context handling and output consistency. A lower per-request cost may not produce savings if an application needs repeated calls, larger prompts or additional validation to obtain an acceptable result.

The release also fits Netlify’s wider effort to bring AI development closer to its deployment platform. The company has opened Agent Runners and AI Gateway to its customer base, as reported in an earlier update, and more recently added the ability for Agent Runners to ask clarifying questions before building apps.

For builders already deploying on Netlify, open model support reduces the need to bolt on another model-management layer purely to control spending. The more important shift is operational: model choice can now become part of a budget-aware workflow, rather than a decision made once when an application is first built.

What is Netlify Open models in the AI Gateway?

Open models in the AI Gateway is an expansion of Netlify’s managed AI access layer. It lets developers use hundreds of open models, including options from DeepSeek, Qwen and Kimi, alongside more expensive frontier models in deployed applications.

How can open models reduce Netlify AI costs?

Netlify says capable open models cost a fraction of frontier-model pricing. Teams can route appropriate automated tasks to these models, helping their available credits last longer without stopping every workflow when the monthly budget becomes constrained.

Are the open models available now?

Netlify’s announcement says the models are now available through AI Gateway. The supplied sources do not specify a separate launch schedule, plan requirement or model-level pricing table, so teams should check Netlify’s current documentation and account controls before changing production workloads.