Intelligence index
Outperforms 67% of compared models
Grok 4.3
Grok 4.3 is a reasoning model from SpaceXAI. It accepts text and image inputs with text output, and is suited for agentic workflows, instruction-following tasks, and applications requiring high factual accuracy. Reasoning can be configured between none/low/medium/high (default low) effort levels.
It supports a 1 million token context window with no output token limit, making it well-suited for long-document analysis, deep research, and multi-step agentic tasks. Pricing is tiered: requests exceeding 200k total tokens are billed at a higher rate.
- Input / output price
- $1.25/$2.5per 1M tokens
- Context
- 1.0M
- Released
- 2026-04-30
Pricing
Artificial Analysis
Explore rankingsCoding index
Outperforms 41% of compared models
Agentic index
Outperforms 45% of compared models
Design Arena
Elo scores and win rates across design and agent tasks.
Category performance
1,000-1,500 EloModel Arena
Asciiart
1123Elo
Win rate 42.5%
UI Component
1188Elo
Win rate 44.9%
Code Categories
1185Elo
Win rate 43.6%
Data Visualization
1181Elo
Win rate 45.1%
Game Development
1171Elo
Win rate 42.8%
Website
1194Elo
Win rate 44%
SVG
1079Elo
Win rate 36.9%
3D
1121Elo
Win rate 40.8%
Agent Arena
Pptxslides
1071Elo
Win rate 32.5%
Agentichtmlslides
1066Elo
Win rate 31.8%
Agenticslides
1072Elo
Win rate 31.9%
Agenticslides(html)
1066Elo
Win rate 31.7%
Agenticslides(python-pptx)
1068Elo
Win rate 32.4%
Blender3d
816Elo
Win rate 12.5%
Agenticgamedev
972Elo
Win rate 28.7%
Htmlslides
905Elo
Win rate 28.4%
Spreadsheet
927Elo
Win rate 34.5%
Python-pptxslides
1072Elo
Win rate 30.8%
Webapps
1126Elo
Win rate 42.8%
Godotgamedev
914Elo
Win rate 20.7%
Androidnative
962Elo
Win rate 22.7%
Fullstack
1006Elo
Win rate 28.1%
Mobile Apps
1058Elo
Win rate 33.4%
Related Models
Grok 4.7
x-aiGrok 4.7 is SpaceXAI's flagship model for coding, agentic tasks, and knowledge work, succeeding Grok 4.6. It is particularly strong at long-running software engineering tasks, verifying its own work, and managing long context, and it improves on its predecessor at professional knowledge work such as drafting documents and presentations. The model was trained with a longer reinforcement learning run weighted toward problems that take many hours to complete, and natively understands the Grok Bot harness for conversational tasks. It ships with a new safeguard stack that pairs strong jailbreak resistance with low refusal rates for legitimate cybersecurity and biology work. SpaceXAI's reported benchmark results use the xhigh reasoning effort.
- Context
- 500K
- Max output
- 450K
- Released
- 2026-09-21
- Input
- $2/M
- Output
- $6/M
- Cache read
- $0.5/M
Grok 4.6
x-aiGrok 4.6 is a model from SpaceXAI with frontier performance on coding, knowledge work, and STEM. It is succeeded by Grok 4.7.
- Context
- 500K
- Max output
- 450K
- Released
- 2026-08-12
- Input
- $2/M
- Output
- $6/M
- Cache read
- $0.5/M
Grok 4.5
x-aiGrok 4.5 is a model from SpaceXAI with frontier performance on coding, knowledge work, and STEM.
- Context
- 500K
- Max output
- 450K
- Released
- 2026-07-08
- Input
- $2/M
- Output
- $6/M
- Cache read
- $0.3/M