Grok 4.3
used in
0
stacks
xAI's flagship API model, billed as its most intelligent and fastest, with strong agentic tool calling.
Grok 4.3 (model ID grok-4.3) is xAI's flagship, described in its docs as "the most intelligent and fastest model we've built." It offers a 1M-token context window at $1.25 per million input tokens and $2.50 per million output tokens, and xAI reports it tops leaderboards in agentic tool calling and instruction following. The API roster also includes Grok 4.20 reasoning/non-reasoning and multi-agent variants at the same price point; several legacy models including grok-3 have been retired from the lineup.
Strengths
- Aggressive pricing ($1.25 input / $2.50 output per million tokens) for a 1M-context flagship
- Vendor-reported leadership in agentic tool calling and instruction following
- Companion multi-agent variant (grok-4.20-multi-agent) targeted at agentic workloads
Tradeoffs
- Stated knowledge cutoff of November 2024 is older than competing flagships
- Rapid model churn; xAI retired eight legacy models (including grok-3) in May 2026, so pinned integrations need active migration