What is New API?
New API is an open-source, self-hosted AI gateway for applications, agents, and teams. New API connects upstream model services, exposes consistent client-facing APIs, and centralizes routing, access management, usage analytics, and cost accounting. The project supports private multi-model services and shared, authorized access to model providers.
New API lists upstream services including OpenAI, Anthropic, Google Gemini, Azure OpenAI, AWS Bedrock, Vertex AI, DeepSeek, and Qwen. New API also supports other compatible services. The repository describes lawful gateway operation, organization-level authentication, multi-model management, and private deployment as intended uses.
What New API can do
New API combines model access, routing controls, account management, and usage accounting in a web console. New API exposes OpenAI Chat Completions and Responses, Anthropic Messages, and Gemini interfaces. Streaming, tool use, reasoning, and multimodal inputs are available where the selected model and connection support them.
- New API configures model mappings, channel priorities, routing weights, retries, and channel affinity.
- New API manages quotas, subscriptions, usage logs, cache accounting, and expression-based pricing.
- New API controls access through users, groups, permissions, and API key restrictions.
- New API supports OAuth/OIDC, passkeys, two-factor authentication, and login session management.
- New API extends asynchronous task APIs through JavaScript plugins with status and output retrieval.
- New API provides a playground, usage inspection, and audit logs in its web console.
How New API works
New API offers a local Docker quick start using SQLite and a persistent data directory. After opening the local console, users complete a setup wizard to create an administrator account. Users then add a channel containing an upstream API key, available models, and a group assignment, and run a channel test.
New API requires configured model pricing and a user quota or valid subscription before the documented first request. Users create a gateway API key with matching group and model access, then configure their client's base URL and credentials. Docker Compose deployment uses PostgreSQL and Redis by default, with additional database examples provided.
Who New API is for
New API serves application developers, agent builders, and teams that need centralized access to multiple model services. New API supports organizations sharing authorized credentials through controlled gateway access, as well as operators maintaining private model services. The web console brings channel configuration, model testing, and usage inspection into the same workflow.
- New API lets teams share authorized model access through groups and restricted API keys.
- New API lets application operators switch providers without configuring every client again.
- New API supports private multi-model services with usage tracking and cost accounting.
New API pricing and access
New API is available as an open-source project under the AGPL-3.0 license through its public GitHub repository. New API provides source code and container deployment instructions rather than a stated hosted subscription price. Users must lawfully obtain their own upstream API keys, accounts, model services, and interface permissions.
New API strengths and limitations
New API brings provider access, routing, authentication, and accounting into a self-hosted gateway. New API also offers a multilingual console and deployment choices covering SQLite, MySQL, and PostgreSQL. These capabilities support private operations, but production deployment involves HTTPS configuration, database persistence, backups, and coordinated secrets across application nodes.
- New API features depend on the channel, upstream model, and conversion path.
- New API protocol-specific tools and fields may not map exactly during conversion.
- New API WebSocket connections require compatible upstream services and channel configuration.
- New API public service operators must meet applicable regulatory and upstream authorization obligations.
New API FAQ
What is New API?
New API is an open-source, self-hosted AI gateway for applications, agents, and teams. New API connects upstream model services and provides consistent APIs alongside model routing, access controls, usage tracking, and cost accounting.
Is New API free?
New API is available as an open-source project under the AGPL-3.0 license, with source code and self-hosting instructions in its public repository. New API requires users to obtain authorized upstream API keys, accounts, model services, and permissions.
Which model providers does New API support?
New API lists OpenAI, Anthropic, Google Gemini, Azure OpenAI, AWS Bedrock, Vertex AI, DeepSeek, Qwen, and other compatible services as upstream options. New API feature availability depends on the selected channel, upstream model, and protocol conversion path.
How do you deploy New API?
New API provides Docker instructions for a local instance using SQLite and Docker Compose instructions using PostgreSQL and Redis by default. New API also documents MySQL, an optional separate log database, persistent storage, HTTPS configuration, and shared secrets for multi-node deployments.
Can New API manage team access and model costs?
New API manages users, groups, fine-grained permissions, and API key restrictions for authorized model access. New API also supports quotas, subscriptions, usage logs, cache accounting, and expression-based pricing for different usage tiers.
Does New API support image, video, and audio APIs?
New API lists image generation, image editing, speech, transcription, and translation endpoints, plus JavaScript plugins for asynchronous image, video, and other tasks. New API capabilities depend on the upstream model, channel configuration, and conversion path.
From the community
No reviews yet
Be the first to review this tool. What you write helps the next person choose.




