Switch between GPT-5, Claude Opus, Gemini 3, Llama, and 200+ more — your memory follows seamlessly.
// Monday: Use GPT-5.2
model: "gpt-5.2"
// Tuesday: Try Claude
model: "anthropic/claude-opus-4.5"
// Wednesday: Test Gemini
model: "google/gemini-3-pro"
// Memory persists across all of them ✓No lock-in. No migration. Just change the model parameter.
OpenAI's flagship model. Improved personality, creativity, and reasoning.
Strong general-purpose model. Excellent balance of capability and cost.
Large pre-trained model for Pro users. Deep knowledge and nuance.
Most powerful reasoning model. SOTA on coding, math, and science benchmarks.
Premium reasoning with extended compute. Best for the hardest problems.
Advanced reasoning model. Thinks before responding for complex tasks.
Most capable Claude. Best for coding, agents, and complex reasoning.
Best balance of speed, intelligence, and cost. Great for everyday tasks.
Fastest and cheapest Claude. Lightning-fast for simple tasks at scale.
Most advanced multimodal model. Perfect for analyzing entire codebases or documents.
Fast and scalable with massive context. Great for real-time apps.
Strong predecessor with native tool use and agentic capabilities.
Fast and cost-efficient. Excellent for high-volume multimodal workloads.
41B/675B MoE flagship. World's best open-weight multimodal model. Apache 2.0 license.
Best-in-class edge model. Vision + reasoning in 14B params. Perfect for local deployment.
24B enterprise-ready compact model. Vision capabilities with Apache 2.0 license.
Cohere, Perplexity, Together, and dozens more. If it's on OpenRouter, it works with MemoryRouter.
One line change. Same code. Persistent memory across every model.