Grok 4 vs Qwen3.6-35B-A3B

Change models
IndexAll categories, out of 100.
49.8 ±4.0
49.8 ±3.1
GapIndex points below the top.
Top
Top
CoverageIndex weight that has results.
95%
95%
SpeedOutput tokens per second.
78/s best
62/s
EloLMArena rating and rank.
1411
—
ReleasedMonth of release.
2025-07
2026-04
Input / 1MUS dollars per 1M input tokens.
—
$0.10
Output / 1MUS dollars per 1M output tokens.
—
$0.90
1K chats2K tokens in, 500 out, each.
—
$0.65
1K documents20K tokens in, 1K out, each.
—
$2.90
1K agent tasks200K tokens in, 20K out, each.
—
$38.00
ContextMaximum tokens in one request.
128K
262K best
AgenticMulti-step tasks with tools.
48.6
51.7 best
CodingCode writing and repair.
50.4 best
48.9
ReasoningLogic problems and puzzles.
46.8
49.0 best
MultimodalTasks with images and text.
49.1
50.8 best
KnowledgeFacts and expert knowledge.
55.5 best
49.4
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
39.1
48.7 best
MathMath problems.
50.0 best
48.3

Benchmarks

10 shared
Only Grok 4: 30Only Qwen3.6-35B-A3B: 43
Grok 4
Qwen3.6-35B-A3B
87.0% best68 / 220
84.8%75 / 220
26.7% best92 / 216
22.2%102 / 216
68.0%120 / 211
71.7% best101 / 211
2.0% best102 / 207
0.3%132 / 207
84.0% best80 / 203
68.9%105 / 203
87.7% best52 / 193
84.1%81 / 193
74.9%72 / 143
95.3% best19 / 143
53.7%69 / 140
64.4% best54 / 140
31.6% best81 / 120
21.5%108 / 120
68.8%76 / 109
75.0% best51 / 109