Grok 4.20 vs Ripple-65B Preview

Change models
Ripple-65B PreviewRicursive Intelligence
IndexAll categories, out of 100.
52.0 ±4.8 best
51.7 ±5.4
GapIndex points below the top.
Top
−0.3 · no clear gap
CoverageIndex weight that has results.
90% best
85%
SpeedOutput tokens per second.
114/s best
76/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-03
2026-11
Input / 1MUS dollars per 1M input tokens.
$1.25
$0.10 best
Output / 1MUS dollars per 1M output tokens.
$2.50
$0.30 best
1K chats2K tokens in, 500 out, each.
$3.75
$0.35 best
1K documents20K tokens in, 1K out, each.
$27.50
$2.30 best
1K agent tasks200K tokens in, 20K out, each.
$300.00
$26.00 best
ContextMaximum tokens in one request.
2M best
1.05M
AgenticMulti-step tasks with tools.
49.9 best
48.7
CodingCode writing and repair.
54.7 best
54.6
ReasoningLogic problems and puzzles.
55.3
58.0 best
MultimodalTasks with images and text.
45.2
—
KnowledgeFacts and expert knowledge.
54.7 best
47.8
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
61.9
MathMath problems.
56.6 best
52.4

Benchmarks

3 shared
Only Grok 4.20: 19Only Ripple-65B Preview: 14
Grok 4.20
Ripple-65B Preview
76.7% best40 / 114
76.5%41 / 114
88.5% best45 / 87
85.2%61 / 87
51.8%65 / 83
61.4% best29 / 83