Loading…
10 models tracked from xAI.
xAI's current flagship Grok model: a reasoning-first multimodal model with a 1M-token context window, text and image input, and aggressively low pricing. Released to the public API on 2026-04-30.
Lighter, faster, lower-cost sibling of the Grok 4.3 flagship. Keeps reasoning and tool use with a 256K-token context window and text+image input, tuned for high-throughput agentic and chat workloads at a fraction of the flagship price.
Latency-optimized Grok 4 variant built for real-time and high-volume workloads. Delivers very high output throughput and low time-to-first-token at the lowest price in the Grok 4 line, with a 256K-token context window and text+image input.
Prior-generation Grok 4 reasoning model, now legacy after the Grok 4.3 release. A multimodal reasoning model with a 1M-token context window and text+image input, still available via the xAI API at slightly higher pricing than the current flagship.
Early Grok 4 generation reasoning model, now legacy. A multimodal model with a 256K-token context window and text+image input, retained in the xAI API for compatibility and reproducibility at legacy pricing.
The original Grok 4 model that launched the Grok 4 generation, now legacy. A multimodal reasoning model with a 256K-token context window and text+image input, kept available in the xAI API for backward compatibility.
Prior-generation Grok 3 model, now legacy. A general-purpose multimodal model with a 131K-token context window and text+image input, retained in the xAI API at low legacy pricing for cost-sensitive and compatibility use cases.
Lower-cost Grok model in the current API lineup with a 256K-token context window and text+image input, priced below the flagship at $1.00/$2.00 per million tokens.
Prior-generation high-performance Grok model with reasoning and non-reasoning variants, 1M-token context, and text+image input. Still listed and offered in the xAI API alongside the newer flagship Grok 4.3.
Latest tracked developments from xAI.