Universal AI Providers & Key Vault

💡 Optimal 2-Phase Strategy: Free Tier First → Final Dev with Gemini / Codex

Start completely free with zero token cost. Once free quotas exhaust, top up directly at your provider (no platform commissions):

Phase 1: Free Tier First ($0 Cost)

Use Google AI Studio (Gemini Free), DeepSeek Free, Zhipu GLM Free, or OpenRouter Free Models for unlimited daily drafting & agent tasking.

Phase 2: Final Production & Code

Switch to Google Gemini 2.5 Pro, OpenAI Codex / GPT-5, or Claude 3.5 Sonnet for high-assurance code generation, complex security analysis, and deployments.

Google Gemini Free Tier Final Dev

Recommended for all workflows. Generous free tier via Google AI Studio.

Get API Key ↗ Top Up / Billing ↗

OpenAI (Codex / GPT-5 / GPT-4o) Final Dev

Industry benchmark for code architecture, deep refactoring, and tool execution.

Get API Key ↗ Top Up / Billing ↗

⚡

OpenRouter (Multi-Model Hub) Free Models

Single key for 200+ models including free Llama 3.3, Qwen 2.5, and DeepSeek.

Get API Key ↗ Add Credits ↗

DeepSeek 深度求索 (V3 / R1) China Ultra-Low Cost

World-class mathematical, coding, and logical reasoning at ultra-low price.

获取 Key ↗ 充值 Top Up ↗

智

Zhipu AI 智谱清言 (GLM-4 / GLM-5) China Free Tier

BigModel open platform with bilingual proficiency and long-context capabilities.

获取 API Key ↗ 充值 Top Up ↗

Moonshot AI 月之暗面 (Kimi K3) China Trial Quota

Pioneering 2M+ token ultra-long context for deep document and codebase digestion.

获取 Key ↗ 充值 Top Up ↗

千

Alibaba Cloud DashScope 阿里百炼 (Qwen / 通义千问) China Free Quota

Qwen 2.5 Coder & Max series across multimodal, code, and reasoning tasks.

开通 / 获取 Key ↗ 费用中心 ↗

Anthropic Claude (3.5 Sonnet / Opus) Final Dev

Exceptional nuanced synthesis, deep document ingestion, and code comprehension.

Get API Key ↗ Billing ↗

𝕏

xAI (Grok / Grok 2 / Grok Build)

Real-time web intelligence and high-throughput responses.

Get API Key ↗ Billing ↗

豆

ByteDance Volcano Engine 火山引擎 (Doubao / 豆包) China

ByteDance's high-speed multimodal and conversational foundation models.

获取 API Key ↗ 充值中心 ↗

言

Baidu Qianfan 百度千帆 (ERNIE / 文心大模型) China

ERNIE 4.0 / Speed enterprise Chinese natural language platform.

获取应用密钥 ↗ 财务充值 ↗

混

Tencent Hunyuan 腾讯混元大模型 China

Tencent Cloud's proprietary multi-domain LLM platform.

管理密钥 ↗ 账号充值 ↗

MiniMax 稀宇科技 (ABAB / Hailuo) China

Advanced text-to-speech, video, and high-context MoA reasoning models.

获取接口密钥 ↗ 账户充值 ↗

阶

StepFun 阶跃星辰 (Step-2 / Step-3.7) China

Step series multimodal and high-speed analytical engines.

开放平台 ↗ 充值中心 ↗

01.AI 零一万物 (Yi-Large / Yi-Lightning) China

Dr. Kai-Fu Lee's high-efficiency open and proprietary model ecosystem.

管理 Key ↗ 财务充值 ↗

百

Baichuan AI 百川智能 (Baichuan 4) China

Specialized medical, knowledge retrieval, and enterprise Chinese reasoning.

获取 Key ↗ 充值中心 ↗

汤

SenseTime SenseNova 商汤日日新 (SenseChat) China

Large scale vision-language and code intelligence.

开放平台 ↗

⚡

Groq LPU (Ultra-Fast Inference)

500+ tokens/sec lightning-speed open models (Llama 3, Mixtral).

Get API Key ↗ Billing ↗

Mistral AI (Large / Codestral / Pixtral)

European frontier models with strong multilingual and coding capabilities.

Get API Key ↗ Billing ↗

Together AI (Cloud Inference)

High-speed fine-tuning and inference for open-weight models.

Get API Key ↗ Billing ↗

Perplexity AI (Sonar / Online Search)

Real-time cited web search intelligence models.

Get API Key ↗

Cohere (Command R+ / Embeddings)

Enterprise RAG, semantic reranking, and multilingual search.

Get API Key ↗ Billing ↗

🔌

Custom AI Provider / Local Endpoint Offline / Local

Connect local Ollama (http://localhost:11434/v1), vLLM, LMStudio, or private enterprise servers.

OpenAI-Compatible Spec