GPT-4.5 vs Grok 3 [Beta]

Change models
GPT-4.5OpenAI
IndexAll categories, out of 100.
38.7 ±8.9
39.1 ±8.5 best
GapIndex points below the top.
−0.4 · no clear gap
Top
CoverageIndex weight that has results.
65% best
60%
SpeedOutput tokens per second.
—
—
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2025-02
2025-02
Input / 1MUS dollars per 1M input tokens.
$75.00
—
Output / 1MUS dollars per 1M output tokens.
$150.00
—
1K chats2K tokens in, 500 out, each.
$225.00
—
1K documents20K tokens in, 1K out, each.
$1,650.00
—
1K agent tasks200K tokens in, 20K out, each.
$18,000.00
—
ContextMaximum tokens in one request.
128K
128K
AgenticMulti-step tasks with tools.
—
—
CodingCode writing and repair.
29.5
34.7 best
ReasoningLogic problems and puzzles.
35.8 best
34.6
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
48.4 best
46.8
MultilingualTasks in many languages.
31.8
—
InstructionTasks with strict prompt rules.
—
—
MathMath problems.
33.8
37.9 best

Benchmarks

8 shared
Only GPT-4.5: 4Only Grok 3 [Beta]: 18
GPT-4.5
Grok 3 [Beta]
68.7%118 / 220
75.8% best106 / 220
37.8%132 / 203
55.6% best119 / 203
69.5%93 / 144
75.4% best79 / 144
0.8% best119 / 137
0.0%126 / 137
10.3% best121 / 132
5.5%125 / 132
0.0%107 / 127
0.0%107 / 127
78.6%29 / 88
88.7% best17 / 88
62.5% best1 / 15
43.6%3 / 15