AI Gateway
AI Gateway is an AI request management and proxy service from Cloudflare, positioned between applications and AI model providers as a unified traffic entry point. Core capabilities include routing, caching, rate limiting, and monitoring of AI API requests, enabling developers to centrally manage calls across multiple AI service providers while offering observability, cost control, and security protection. It is suited for enterprises and development teams deploying AI applications at scale.
Core Facts
Gemini 4 Argon is positioned by Google as a 'workhorse' model built specifically for coding and cybersecurity work.
Typical capabilities of AI gateways include centralized logging, automatic degradation and fallback, per-project cost tracking, and a unified interface.
AI Gateway is the unified entry point for model calls provided by Vercel, converging the APIs of multiple model providers such as OpenAI, Anthropic, Google, GLM, and Kimi into a single endpoint, and it handles routing, retries, caching, and usage statistics.
Timeline (last 90 days)
The prices for Mistral Large 4's input tokens, cached input tokens, and output tokens have all been lowered.
Mistral Large 4 updated its model weights on AI Gateway and lowered its price at the same time, while the model ID `mistral/mistral-large-4` remains unchanged.
Vendors updating weights under the same ID through a gateway is becoming a common practice, at the cost of weaker version traceability, since the same model ID may behave differently at different times.
The code path for calling Mistral Large 4 on AI Gateway has not changed; when using the AI SDK, you still pass `mistral/mistral-large-4` and call `streamText`.
For production systems with extremely high stability requirements, it is advisable to establish evaluation and regression mechanisms at the gateway layer rather than assuming that a model with the same name is always equivalent.
AI Gateway has launched the "Virtual Models" feature, which abstracts model selection and parameter configuration from application code into configuration saved on the gateway side; callers invoke it via a custom model ID (such as vmc/coding-agent), and the underlying model and settings can be modified centrally without redeploying the application or Agent.
For coding Agents, you first need to connect AI Gateway via `vercel ai-gateway setup`, then select the virtual model in the Agent's model selector or configuration.
On the Virtual Models page, click Create Virtual Model, fill in the slug, select the underlying model, configure the required settings, and save to create a virtual model; it can also be created via the Vercel CLI command `vercel ai-gateway virtual-models create`.
After a virtual model's configuration is changed, all applications and Agents that call that ID automatically adopt the new configuration within a few minutes, with no code changes or redeployment needed.
Virtual models are compatible with all model types on AI Gateway, including image, video, speech, embedding and real-time models.
40 more timeline events
All Facts (20)
Gemini 4 Argon is positioned by Google as a 'workhorse' model built specifically for coding and cybersecurity work.
90%VerifiedTypical capabilities of AI gateways include centralized logging, automatic degradation and fallback, per-project cost tracking, and a unified interface.
80%VerifiedAI Gateway is the unified entry point for model calls provided by Vercel, converging the APIs of multiple model providers such as OpenAI, Anthropic, Google, GLM, and Kimi into a single endpoint, and it handles routing, retries, caching, and usage statistics.
75%VerifiedBYOK (Bring Your Own Key) means developers use their own provider API Key to send requests through the gateway, while billing still goes through the direct account between the developer and the provider.
75%VerifiedPrivate deployment keeps model weights and inference runtimes entirely within the enterprise's own infrastructure so that data never leaves the domain, whereas in the cloud API model, text data must be transmitted to third-party servers.
75%VerifiedTraditional mobile AI application architecture requires building an intermediate server to hold the API key and forward requests in order to prevent the key from being exposed through reverse engineering, but this introduces operational costs, latency overhead, and scaling pressure.
75%VerifiedAI Gateway offers a ZDR (Zero Data Retention) feature, whereby request data is not retained by the vendor for training or other purposes.
75%VerifiedZero Data Retention means that after completing a user's request, the service provider does not retain any input or output data on its servers.
75%VerifiedThe AI middleware layer (AI Gateway/AI Infrastructure Layer) includes routing gateways, vector databases (such as Pinecone and Weaviate), RAG frameworks (such as LangChain and LlamaIndex), and observability platforms (such as Langfuse and Helicone).
75%VerifiedVercel AI Gateway is an AI gateway service launched by Vercel, providing features such as caching, rate limit management, and observability.
75%VerifiedThe essence of intelligent routing is to dynamically assign requests to the most suitable model based on the complexity and type of the request.
70%VerifiedOpenRouter is an AI model aggregation and routing platform that allows developers to access hundreds of models from multiple vendors through a unified OpenAI-compatible API interface.
70%VerifiedVercel AI Gateway is a unified API gateway service that allows developers to call models from multiple vendors through a single interface, similar to OpenRouter.
65%VerifiedAI Gateway plays the role of a unified entry point, so teams can access multiple model providers while maintaining a single set of credentials, billing, and logging systems.
65%VerifiedAI Gateway has built-in features such as custom reporting, Zero Data Retention support, API key budget management and routing rules.
65%Verifiedd1's decision calls appear in AI Gateway logs and are counted in budget statistics, allowing reuse of AI Gateway's cost attribution, call auditing, and quota control processes.
65%VerifiedAI Gateway supports features such as custom reports, zero data retention, API key budget management, and routing rules.
65%VerifiedFor coding agent tools such as Claude Code, Codex, and Cursor, you need to first install the latest version of Vercel CLI and run the 'vercel ai-gateway setup' configuration command.
65%VerifiedAI gateways handle cross-cutting concerns related to interactions with large language models, and are responsible for model routing, key and credential management, and cost and quota control.
65%VerifiedThe Gateway in Claude Code acts as an API gateway and serves as the unified entry point for users to configure and manage different AI model interfaces.
65%