AI Providers

OpenAI is an AI research and deployment company. Our mission is to ensure that artificial general intelligence benefits all of humanity.

NovitaAI's OpenAI-compatible large language models

Microsoft Azure - enterprise-grade OpenAI models

DeepInfra inference platform with OpenAI-compatible API for hosting open-source models.

Google AI Studio is a platform for accessing Google's Gemini models.

Google Vertex AI is a platform for accessing Google's Gemini models via Vertex AI.

Alibaba Cloud's Qwen large language models with OpenAI-compatible API

Amazon Bedrock - fully managed service for foundation models

Tencent Cloud's TokenHub model gateway, serving Tencent's own Hunyuan models alongside third-party models through a single OpenAI-compatible API.

Anthropic is a research and deployment company focused on building safe and useful AI.

xAI's Grok large language models

DeepSeek's high-performance language models with OpenAI-compatible API

Together AI is a platform for running large language models in the cloud with fast inference.

Access partner models (e.g. xAI Grok) via Google Cloud Vertex AI's OpenAI-compatible Chat Completions endpoint.

Mistral AI's large language models

EmberCloud provides access to a variety of large language models via an OpenAI-compatible API

Runware provides fast, cost-efficient inference for open and frontier LLMs through an OpenAI-compatible API.

Z AI's OpenAI-compatible large language models

Baidu's Qianfan platform serving DeepSeek, GLM, Kimi, MiMo, and Hy3 models

Access Claude models via Google Cloud Vertex AI with the Anthropic Messages API.

SCX.ai is an Australian sovereign AI platform providing OpenAI-compatible general-purpose inference endpoints for a range of open models and SCX's own models, hosted on renewable-powered infrastructure.

Anthropic Claude models on Microsoft Foundry via the Anthropic Messages API

Cerebras high-performance inference with ultra-fast throughput

CanopyWave is a platform for running large language models with OpenAI-compatible API

SCX.ai is an Australian sovereign AI platform providing OpenAI-compatible Turbo inference endpoints — up to 4x faster than comparable providers — for a range of open models and SCX's own models, hosted on renewable-powered infrastructure.

Fireworks AI serves open-weight models on a fast, OpenAI-compatible inference platform.

OpenAI frontier models on Amazon Bedrock's Mantle and Runtime Responses APIs

Moonshot AI's OpenAI-compatible large language models

Meta's Model API serving Muse reasoning and image models via an OpenAI-compatible API

Xiaomi MiMo API Open Platform provides access to the MiMo series of large language models.

ElevenLabs provides lifelike, low-latency text-to-speech models in 70+ languages.

Microsoft Azure AI Foundry - third-party models (Grok, Llama, Mistral, ...) via the Azure Models inference endpoint

Perplexity's AI models for search and conversation with real-time web access

ByteDance's ModelArk platform with OpenAI-compatible API for large language models

Sakana AI's Fugu multi-agent orchestration models, served through a single OpenAI-compatible API.

Groq's ultra-fast LPU inference with various models

AtlasCloud provides unified APIs for video, image, audio, and language generation models.

MiniMax's large language models with advanced reasoning and coding capabilities

Meta's discounted, training-eligible Model API tier for Muse models

Gonka24 serves open-weight large language models via an OpenAI-compatible inference gateway.

Runpod's serverless public inference endpoints

Inference.net is a platform for running large language models in the cloud.

RanoAI serves open-weight large language models on Furiosa RNGD NPU hardware via an OpenAI-compatible inference API.

Consensus Protocol serves open-weight large language models on dedicated GPU hardware it operates, via an OpenAI-compatible inference API.

Atria ASI serves the Atria Dawn agentic research model through an OpenAI-compatible inference API.

TypeSafe AI serves Jev, a System One decision model that answers typed questions about a state with calibrated probabilities instead of generated text.