Skip to main content

Grok 4.5 vs Grok 4.6

SpecGrok 4.5Grok 4.6
Respan Index46.5 (#52)53.4 (#32)
Blended price per 1M$3$3
Context window500K500K
Input / output per 1M$2 / $6$2 / $6
Max output--
Inputtext, visiontext, vision
Released2026-07-082026-08-12

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryGrok 4.5Grok 4.6
Coding25.5 Grok 4.534.0 Grok 4.6
Agents64.7 Grok 4.565.8 Grok 4.6
Reasoning42.8 Grok 4.559.4 Grok 4.6
Knowledge & language49.7 Grok 4.551.1 Grok 4.6
Human preference55.0 Grok 4.556.5 Grok 4.6

How the Respan Index works

Benchmarks both report

Grok 4.6 leads on 21 of the 26 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkGrok 4.5Grok 4.6
LMArena Elo14661457
LMArena Vision (LMArena)12801264
LMArena WebDev (LMArena)15521618
MirrorCode (Epoch AI)32.2%41.1%
Chess Puzzles (Epoch AI)35%33%
LMArena Agent (LMArena)0.0079-0.006
APEX-Agents (Mercor)56.2%65.3%
DeepSWE v1.1 (Datacurve)53.8%67.5%
FrontierMath Tier 4 v2 (Epoch AI)24.4%31.7%
FrontierMath Tiers 1-3 v2 (Epoch AI)57.2%66%
Furniture Assembly (Epoch AI)22.5%40%
GPQA Diamond93.4%94%
LiveBench Agentic Coding (LiveBench)56.5%57%
LiveBench Coding (LiveBench)68.6%76.8%
LiveBench Data Analysis (LiveBench)73%73.9%
LiveBench Instruction Following (LiveBench)71.5%71.9%
LiveBench Language (LiveBench)82.8%83.7%
LiveBench Mathematics (LiveBench)90.8%92.6%
LiveBench Reasoning (LiveBench)87.2%90.5%
OTIS Mock AIME 2024-2025 (Epoch AI)97.8%99.2%
SAGE (Vals AI)35%28.9%
SimpleQA Verified48.3%49.3%
Terminal-Bench 4.0 (Vals AI)8.6%17.2%
Vending-Bench 2 (Andon Labs)3887.439047.03
WeirdML (Håvard Tveit Ihle)46.4%67.3%
SimpleBench (SimpleBench)70%75.9%

FAQ

Is Grok 4.5 better than Grok 4.6?

Grok 4.6 scores 53.4 on the Respan Index (#32), ahead of Grok 4.5 at 46.5 (#52). Grok 4.6 leads on 21 of the 26 benchmarks both report.

Which is cheaper, Grok 4.5 or Grok 4.6?

Grok 4.5 lists at $2 input and $6 output per 1M tokens; Grok 4.6 at $2 and $6. Both cost about the same per token at list price.

Which has the larger context window?

Both accept 500K tokens of context.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.