Grok 4.3
xAI (SpaceXSI) · Multimodal model
Grok 4.3 is xAI's lower-cost general model, described in xAI docs as fast and reliable with strong tool calling and instruction following. It has a 1M-token context and costs less than half of Grok 4.7 per output token. It is…
Current
ReleasedApr 2026
Context window1,000,000 tokens
Input → outputtext, image → text
API price$1.25 / 1M input, $0.20 / 1M cached input, $2.50 / 1M output (prompts < 200k tokens); $2.50 input, $0.40 cached, $5.00 output (prompts ≥ 200k tokens). Batch API: 20% discount.
Consumer access—
Best for
Cheaper, fast Grok workloads that need reliable tool calling and a 1M-token context.
Overview
Grok 4.3 is xAI's lower-cost general model, described in xAI docs as fast and reliable with strong tool calling and instruction following. It has a 1M-token context and costs less than half of Grok 4.7 per output token. It is a practical pick for high-volume or tool-heavy jobs.
Strengths and limits
- Lowest per-token price among current general Grok models
- 1M-token context window
- Adjustable reasoning from 'none' to 'xhigh'; Batch API supported
- Not a frontier model; well behind Grok 4.7 and rivals on hard tasks
- Text-only output