Grok 4 vs MiniMax M2.5
Change modelsGrok 4xAI
MiniMax M2.5MiniMax
IndexAll categories, out of 100.
50.0 ±4.0
50.2 ±7.1 best
GapIndex points below the top.
−0.2 · no clear gap
Top
CoverageIndex weight that has results.
95% best
80%
SpeedOutput tokens per second.
78/s best
44/s
EloLMArena rating and rank.
1411 best
1359
ReleasedMonth of release.
2025-07
2026-02
Input / 1MUS dollars per 1M input tokens.
—
$0.27
Output / 1MUS dollars per 1M output tokens.
—
$1.08
1K chats2K tokens in, 500 out, each.
—
$1.08
1K documents20K tokens in, 1K out, each.
—
$6.48
1K agent tasks200K tokens in, 20K out, each.
—
$75.60
ContextMaximum tokens in one request.
128K
200K best
AgenticMulti-step tasks with tools.
48.0
51.6 best
CodingCode writing and repair.
50.9 best
50.7
ReasoningLogic problems and puzzles.
47.3
48.8 best
MultimodalTasks with images and text.
49.6
—
KnowledgeFacts and expert knowledge.
55.8 best
48.9
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
39.2
—
MathMath problems.
50.5
54.3 best
Benchmarks
13 sharedOnly Grok 4: 27Only MiniMax M2.5: 7
Grok 4
MiniMax M2.5