Retour aux modèles
xAI
Grok 4
Propriétaire4 variantes
Grok 4
instructxAI's Grok 4 flagship — frontier reasoning and multimodal model with a 256K context window and tool-use capabilities.
256K tokens$3 in · $15 out / 1MTransformerProprietary
Grok 4
55.5%Grok 4.1 Fast
55.5%Grok 4.20
55.5%exo-unoptimized (grok-4-fast + gemini-2.5-flash)
instruct
xAI's ultra-fast Grok 4.1 Fast — 2M context at a fraction of the cost, optimised for high-throughput production.
2M tokens$0.20 in · $0.50 out / 1MTransformer
reasoning
xAI's Grok 4.20 with 2M token context — advanced reasoning and multi-agent capabilities for complex agentic workflows.
2M tokens$2 in · $6 out / 1MTransformer
instruct
xAI's Grok 4.20 Multi-Agent — specialised for orchestrating complex multi-agent workflows with 2M token context.
2M tokens$2 in · $6 out / 1MTransformer
