$ basecode init my-project
> Initializing project with AI orchestration...
> Connected: OpenAI, Anthropic, Gemini, DeepSeek, Groq
> AI agents deployed: CodeWriter, Tester, Reviewer
$ basecode run --agents 3
> [Agent-1] Writing authentication module...
> [Agent-2] Generating unit tests...
> [Agent-3] Running security audit...
> All tasks completed. 0 bugs found. ✓
$ ▊
Unified API, seamless failover, one dashboard
Everything you need to build software with AI at maximum velocity
10+ providers: OpenAI, Anthropic, Gemini, Mistral, DeepSeek, Groq, OpenRouter, Together, Perplexity, Fireworks. Auto failover between them.
Control costs: single provider mode, single model lock, or smart OpenRouter auto-selection with ultra-cheap/balanced/quality tiers.
40+ indexed models with auto-sync. Pricing, context window, capabilities — all tracked and auto-updated when providers add new models.
When one API hits limits, requests auto-transfer to equivalent models from other providers. Priority, round-robin, cost-optimal, and latency strategies.
Custom system prompts per project and provider. 6 presets: coding assistant, reviewer, architect, mentor, tester, documenter.
Model Context Protocol for AI-to-AI communication. Delegate tasks between models, coordinate multi-agent workflows.
Control costs: single provider mode, single model lock, or smart OpenRouter auto-selection with ultra-cheap/balanced/quality tiers.
Single Provider
Use one provider, multiple models
Single Model
Lock to one exact model
OpenRouter Smart
Auto-select by budget tier
// Economy Mode API
PUT /ai/economy-mode/:projectId
{
"mode": "openrouter-auto",
"openRouterBudget": "balanced"
}
// Response:
{ "success": true, "model": "deepseek/deepseek-chat" }
Simple, transparent pricing. Pay only for what you need.