Martian Gateway

Paid APIWeb

What Martian Gateway is

Martian Gateway is a paid, usage-based API for accessing more than 200 AI models from multiple providers through one interface. Its available-models page lists 286 models from 48 providers and shows pricing for each model. Requests identify a model with its full creator/model-name identifier, while the Gateway accepts OpenAI Chat Completions and Anthropic Messages API formats. Developers can connect through the OpenAI SDK, Anthropic SDK, Vercel AI SDK, LiteLLM, direct HTTP requests, or listed coding tools including Cursor, Claude Code, Claude Desktop, Cline, OpenCode, and Codex. API capabilities include streaming, caching, prompt caching, Anthropic message caching, and function calling. The Martian Dashboard tracks real-time usage, model performance, and request history. Authentication uses bearer API keys. Credits are required, fair-usage rate limits apply, and model availability may vary by region.

Who it suits

It suits developers who want one API for choosing models by use case while using familiar SDKs, HTTP clients, or coding tools. Teams needing usage analytics and model-performance views may also find the Dashboard relevant.

What it does well

286 models across 48 providers
OpenAI and Anthropic API compatibility
Streaming, caching, and function calling
Integrations with major SDKs and coding tools
Real-time usage and request history

What to know first

Paid, usage-based fees
Credits required for use
Fair-usage rate limits apply
Model availability may vary by region

Verdict

Martian Gateway combines a broad model catalog with compatibility across common API formats and development tools. Check regional availability, credit requirements, and rate limits before adopting it for production workloads.

Facts from Martian Gateway's own pages

API compatibility: The Gateway supports OpenAI-compatible and Anthropic-compatible APIs and client libraries. · gateway-docs.withmartian.com, 11 Oct 2026
API key security: Martian advises users to keep API keys secret, store them in environment variables or a secrets manager, rotate them regularly, and monitor usage. · gateway-docs.withmartian.com, 11 Oct 2026
Authentication: Requests authenticate with API keys sent as bearer tokens in the Authorization header. · gateway-docs.withmartian.com, 4 Oct 2026
Developer tool integrations: Documented coding integrations include Cursor, Claude Code, Claude Desktop, Cline, OpenCode, and Codex. · gateway-docs.withmartian.com, 4 Oct 2026
Gateway features: The API reference documents streaming, caching, prompt caching, and function calling capabilities. · gateway-docs.withmartian.com, 4 Oct 2026
Integrations: Documented integrations include OpenAI SDK, Anthropic SDK, Vercel AI SDK, HTTP clients, LiteLLM, Cursor, Claude Code, Claude Desktop, Cline, OpenCode, and Codex. · gateway-docs.withmartian.com, 11 Oct 2026
Intended users: Martian describes Gateway as a single API for choosing models by use case while staying compatible with popular frameworks. · gateway-docs.withmartian.com, 11 Oct 2026
Key security guidance: Martian advises keeping API keys secret, storing them in environment variables or a secrets manager, rotating them, and monitoring usage. · gateway-docs.withmartian.com, 4 Oct 2026
Maker: Martian describes itself as an AI research lab focused on understanding machine intelligence. · withmartian.com, 4 Oct 2026
Model availability: The support page says models may not be available in all regions. · gateway-docs.withmartian.com, 11 Oct 2026
Model catalog: The Available Models page lists 286 models across 48 providers and says prices are shown per 1 million tokens for input and output. · gateway-docs.withmartian.com, 11 Oct 2026
Model selection: API requests select a model using its full creator/model-name identifier. · gateway-docs.withmartian.com, 4 Oct 2026
Pricing: Martian’s terms say Gateway fees are usage-based and set at the then-current rates on its pricing page. · withmartian.com, 11 Oct 2026
Privacy and customer data: Martian’s privacy policy says it does not cover Customer Data processed on behalf of customers under a data processing addendum; customer contracts govern that processing. · withmartian.com, 11 Oct 2026
Rate limits: The support page says the Gateway applies fair-usage rate limiting and users can contact support to request higher rate limits. · gateway-docs.withmartian.com, 11 Oct 2026
Request limit: The Chat Completions endpoint does not support the n parameter for generating multiple responses from a single prompt. · gateway-docs.withmartian.com, 11 Oct 2026
SDK integrations: Documented SDK options include OpenAI, Anthropic, Vercel AI SDK, and LiteLLM, alongside direct HTTP requests. · gateway-docs.withmartian.com, 4 Oct 2026
Streaming and caching: The Gateway supports streaming responses and Anthropic message caching for its Messages endpoint. · gateway-docs.withmartian.com, 11 Oct 2026
Support: Gateway integration support is available by email at [email protected]. · gateway-docs.withmartian.com, 11 Oct 2026
Usage charges: The documentation describes insufficient-credit errors and directs users to add credits through the Martian Dashboard. · gateway-docs.withmartian.com, 4 Oct 2026
Usage dashboard: The Martian Dashboard provides real-time usage, model performance, and request history. · gateway-docs.withmartian.com, 11 Oct 2026
What it does: Martian Gateway provides unified API access to more than 200 AI models from multiple providers. · gateway-docs.withmartian.com, 11 Oct 2026