Grok 4.3 vs Grok 4.5

xAI against xAI, the same lab's consecutive models. Prices, scores and speed below are live.

  • Grok 4.3 costs 48% less on blended price ($2.34 against $4.50 per 1M tokens, 3 input to 1 output).
  • Grok 4.5 scores higher on the Artificial Analysis Intelligence Index: 39.1 against 25.4.
  • On the 38 benchmarks both have scores on, Grok 4.3 leads on 1 and Grok 4.5 on 37.
  • Grok 4.3 reads more context: 1M tokens against 500K.
  • Grok 4.5 is newer: listed Jul 8, 2026, against Apr 30, 2026.

Side by side

Grok 4.3Grok 4.5
LabxAIxAI
ListedApr 30, 2026Jul 8, 2026
Input per 1M tokens$1.88$3.00
Output per 1M tokens$3.75$9.00
Cached input per 1M$0.30$0.45
Blended (3:1) per 1M$2.34$4.50
Context window1M500K
Max output900K450K
Knowledge cutoff--
Intelligence index25.439.1
Coding index42.272.4
ReasoningYesYes
Reads imagesYesYes
Tool useYesYes
Structured outputYesYes

Every benchmark both are scored on

Each model's best score per benchmark, with its source. Scores from different sources or settings are not always comparable; each links to where it was published.

BenchmarkGrok 4.3Grok 4.5
AIME93.3%Epoch AI97.8%Epoch AI
Artificial Analysis Agentic Index15.5Artificial Analysis and OpenRouter41.2Artificial Analysis and OpenRouter
Artificial Analysis Coding Index42.2Artificial Analysis and OpenRouter72.4Artificial Analysis and OpenRouter
Artificial Analysis Intelligence Index25.4Artificial Analysis and OpenRouter39.1Artificial Analysis and OpenRouter
Chess Puzzles18%Epoch AI35%Epoch AI
Design Arena: 3D1116 EloDesign Arena1247 EloDesign Arena
Design Arena: Android Native (agentic)963 EloDesign Arena1235 EloDesign Arena
Design Arena: ASCII Art1124 EloDesign Arena1255 EloDesign Arena
Design Arena: Code1182 EloDesign Arena1270 EloDesign Arena
Design Arena: Data Viz1177 EloDesign Arena1265 EloDesign Arena
Design Arena: Full-stack (agentic)993 EloDesign Arena1233 EloDesign Arena
Design Arena: Game Dev1167 EloDesign Arena1258 EloDesign Arena
Design Arena: Game Dev (agentic)955 EloDesign Arena1182 EloDesign Arena
Design Arena: Godot Game Dev (agentic)929 EloDesign Arena1229 EloDesign Arena
Design Arena: HTML Slides (agentic)905 EloDesign Arena1101 EloDesign Arena
Design Arena: Mobile Apps (agentic)1056 EloDesign Arena1213 EloDesign Arena
Design Arena: PPTX Slides (agentic)1072 EloDesign Arena1189 EloDesign Arena
Design Arena: SVG1078 EloDesign Arena1206 EloDesign Arena
Design Arena: UI Component1188 EloDesign Arena1275 EloDesign Arena
Design Arena: Web Apps (agentic)1113 EloDesign Arena1204 EloDesign Arena
Design Arena: Website1193 EloDesign Arena1277 EloDesign Arena
DTBench90.7%Epoch AI96.5%Epoch AI
FrontierMath Tier 4 (v2)14.6%Epoch AI24.4%Epoch AI
FrontierMath Tiers 1-3 (v2)42.8%Epoch AI57.2%Epoch AI
GPQA Diamond88.8%Epoch AI93.4%Epoch AI
LMArena (Coding)1414 EloLMArena1474 EloLMArena
LMArena (Overall)1397 EloLMArena1448 EloLMArena
LMArena: Creative Writing1380 EloLMArena1442 EloLMArena
LMArena: Hard Prompts1396 EloLMArena1462 EloLMArena
LMArena: Longer Query1393 EloLMArena1463 EloLMArena
LMArena: Math1396 EloLMArena1479 EloLMArena
LMArena: Multi-Turn1406 EloLMArena1455 EloLMArena
LMCA38.3Epoch AI45.2Epoch AI
ProofBench11%Epoch AI31%Epoch AI
SciCode48.3%Epoch AI55%Epoch AI
SimpleQA Verified33.2%Epoch AI48.3%Epoch AI
WebDev Arena1357 EloEpoch AI1555 EloEpoch AI
WeirdML49.9%Epoch AI46.4%Epoch AI

More comparisons

Where these figures come from

Prices are GenMagic's own, what the studio and the API charge, recomputed from each provider's price list every few minutes. The indices are the Artificial Analysis indices, and every benchmark score links to its source, re-read daily. Measured speed is GenMagic's own generation records, in aggregate. See xAI models, xAI models, every LLM's price and our sources.