Celeris-1 vs Granite 4.2 3B

Change models
Celeris-1Celeris
IndexAll categories, out of 100.
33.7 ±8.2 best
33.3 ±5.8
GapIndex points below the top.
Top
−0.4 · no clear gap
CoverageIndex weight that has results.
75%
85% best
SpeedOutput tokens per second.
722/s
—
EloLMArena rating and rank.
—
1293
ReleasedMonth of release.
2026-07
2026-08
Input / 1MUS dollars per 1M input tokens.
$0.20
—
Output / 1MUS dollars per 1M output tokens.
$0.70
—
1K chats2K tokens in, 500 out, each.
$0.75
—
1K documents20K tokens in, 1K out, each.
$4.70
—
1K agent tasks200K tokens in, 20K out, each.
$54.00
—
ContextMaximum tokens in one request.
128K
128K
AgenticMulti-step tasks with tools.
33.8 best
33.2
CodingCode writing and repair.
24.4
34.6 best
ReasoningLogic problems and puzzles.
35.0 best
30.5
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
35.3 best
28.4
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
26.2
57.6 best
MathMath problems.
—
35.1

Benchmarks

7 shared
Only Celeris-1: 3Only Granite 4.2 3B: 10
Celeris-1
Granite 4.2 3B
6.8% best158 / 216
6.6%160 / 216
38.0% best163 / 211
24.3%184 / 211
0.0%145 / 207
0.0%145 / 207
63.1% best144 / 193
55.9%160 / 193
0.0%100 / 132
0.0%100 / 132
21.6%111 / 113
25.3% best107 / 113
75.9% best45 / 73
67.8%58 / 73