GPT-4o Mini
standardFast, cost-effective
GPT-4o
premiumMultimodal flagship
GPT-5.6 Luna
standardEfficient high-volume GPT-5.6
GPT-5.6 Terra
premiumBalanced intelligence and cost
GPT-5.6 Sol
frontierFrontier GPT-5.6 for complex work
GPT-5.6
frontierAlias for GPT-5.6 Sol
o3-mini
premiumReasoning optimized
Claude Haiku 4.5
standardFast Claude responses
Claude Sonnet 4
premiumBalanced Claude 4
Claude Sonnet 5
premiumLatest Sonnet mid-tier
Claude Opus 4
frontierClaude 4 maximum capability
Claude Opus 5
frontierLatest Opus frontier
Gemini 2.5 Flash
standardUltra-fast Gemini
Gemini 3.6 Flash
standardLatest Gemini Flash workhorse
Gemini 2.5 Pro
premiumGoogle multimodal Pro
Gemini 3.1 Pro
frontierLatest Gemini Pro preview
DeepSeek V3
premiumStrong coding and math
DeepSeek R1
premiumChain-of-thought reasoning
Mistral Large
premiumEuropean flagship
Mistral Small
standardEfficient general purpose
Codestral
premiumCode-specialized
Qwen Max
premiumAlibaba flagship
Groq Llama 70B
standardUltra-low latency
Grok 2
premiumxAI conversational model
Grok 4.3
premiumLong-context Grok workhorse
Grok 4.5
frontierLatest Grok coding and agents flagship
OpenRouter Auto
standardRoutes to best available
Cerebras Llama 3.1 8B
standardUltra-fast Cerebras inference
Cerebras Llama 3.3 70B
premiumFast 70B on Cerebras
SambaNova Llama 405B
frontierLarge model on SambaNova Cloud
SambaNova DeepSeek V3
premiumDeepSeek on SambaNova
Fireworks Llama 70B
standardOptimized inference
Kimi K2.6
standardGeneral-purpose Kimi with vision and tools
Kimi K2.7 Code
premiumKimi coding agent model
Kimi K3
frontierMoonshot flagship — 1M context, deep reasoning
Claude Sonnet (OpenRouter)
premiumClaude Sonnet via OpenRouter
Qwen 2.5 72B (OpenRouter)
standardAlibaba Qwen via OpenRouter
Groq Mixtral
standardFast mixture-of-experts