TheRouter integrates with 6 AI model providers: Anthropic (Claude series), OpenAI (GPT and o-series), Google (Gemini series), xAI (Grok series), AWS Bedrock (Nova and Titan series), and SiliconFlow (200+ models including 3 free models). Each provider has dedicated provider services on our infrastructure for direct API access.

Providers

AI Model Providers

TheRouter routes your requests directly to leading AI model providers — no pools, no middlemen, no substitution. Every request hits the real model at the real provider.

Anthropic

US
9 models

Claude — frontier intelligence for coding, agents, and enterprise

ReasoningCodingVisionTool use
View profile →

OpenAI

US
51 models

GPT & o-series — the broadest model portfolio in the industry

ReasoningVisionAudioEmbeddings
View profile →

Google

US
17 models

Gemini — multimodal intelligence at 1M+ token context

MultimodalLong contextVisionAudio
View profile →

xAI

US
6 models

Grok — real-time reasoning with X platform integration

ReasoningVisionFast inferenceTool use
View profile →

AWS Bedrock

US (us-east-2)
6 models

Nova & Titan — enterprise AI with AWS compliance guarantees

MultimodalLong contextEmbeddingsEnterprise
View profile →

Mistral

EU (France)
11 models

European AI excellence — from 3B edge models to 675B frontier reasoning

ReasoningCodingVisionMultilingual
View profile →

Groq

US
4 models

LPU-powered inference — the fastest AI inference in the world

Ultra-fast inferenceOpen source modelsTool useLow latency
View profile →

DeepSeek

China / Global
4 models

Open-source frontier models with 1M context and industry-leading coding performance

ReasoningCodingLong contextOpen source
View profile →

SiliconFlow

Global
21 models

200+ models including 3 completely free — DeepSeek, Qwen & more

Free modelsReasoningCodingVision
View profile →

MiniMax

Global
7 models

Agent-native frontier models with 200K context and interleaved reasoning

ReasoningCodingTool useAgents
View profile →

Zhipu

China
29 models

GLM series — China-direct supply with vision, OCR, image, video, and embeddings

VisionOCRReasoningImage generation
View profile →

Doubao (Volcengine Ark)

China
17 models

Doubao series — ByteDance's full-stack model platform (chat, vision, reasoning, embedding, image, video)

ReasoningVisionEmbeddingsImage generation
View profile →

Aliyun Bailian (Qwen)

China + Singapore
13 models

Qwen + Wan — Alibaba's full Qwen3 lineup plus Wan image generation, dual-region (CN + Singapore)

ReasoningCodingVisionLong context
View profile →
Customer Support