Grok Build 0.1 vs Qwen3.6-27B

Change models
IndexAll categories, out of 100.
51.5 ±7.2 best
51.3 ±2.4
GapIndex points below the top.
Top
−0.2 · no clear gap
CoverageIndex weight that has results.
85%
100% best
SpeedOutput tokens per second.
—
44/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-05
2026-04
Input / 1MUS dollars per 1M input tokens.
$1.00
$0.32 best
Output / 1MUS dollars per 1M output tokens.
$2.00 best
$3.20
1K chats2K tokens in, 500 out, each.
$3.00
$2.24 best
1K documents20K tokens in, 1K out, each.
$22.00
$9.60 best
1K agent tasks200K tokens in, 20K out, each.
$240.00
$128.00 best
ContextMaximum tokens in one request.
256K
262K best
AgenticMulti-step tasks with tools.
52.5
55.1 best
CodingCode writing and repair.
46.6
52.1 best
ReasoningLogic problems and puzzles.
56.2 best
48.5
MultimodalTasks with images and text.
—
52.4
KnowledgeFacts and expert knowledge.
55.9 best
48.0
MultilingualTasks in many languages.
—
60.6
InstructionTasks with strict prompt rules.
57.8 best
43.0
MathMath problems.
48.8
50.2 best

Benchmarks

9 shared
Only Qwen3.6-27B: 55
Grok Build 0.1
Qwen3.6-27B
13.3% best108 / 130
11.9%110 / 130
27.2%97 / 120
32.4% best80 / 120
70.8% best53 / 80
70.4%55 / 80
76.4% best67 / 80
70.3%76 / 80
45.8% best51 / 66
39.3%63 / 66
65.4%65 / 66
71.8% best55 / 66
65.2% best41 / 66
53.2%65 / 66
72.5% best57 / 66
63.3%62 / 66
78.4%62 / 66
79.9% best58 / 66