Gemini 3.1 Pro Preview vs GPT-5.3 Codex

Change models
IndexAll categories, out of 100.
61.5 ±3.6
62.0 ±5.3 best
GapIndex points below the top.
−0.5 · no clear gap
Top
CoverageIndex weight that has results.
95% best
85%
SpeedOutput tokens per second.
77/s best
52/s
EloLMArena rating and rank.
1480
—
ReleasedMonth of release.
2026-02
2026-02
Input / 1MUS dollars per 1M input tokens.
$2.00
$1.75 best
Output / 1MUS dollars per 1M output tokens.
$12.00 best
$14.00
Batch output / 1MBatch price per 1M output tokens.
$6.00
—
1K chats2K tokens in, 500 out, each.
$10.00 best
$10.50
1K documents20K tokens in, 1K out, each.
$52.00
$49.00 best
1K agent tasks200K tokens in, 20K out, each.
$640.00
$630.00 best
ContextMaximum tokens in one request.
1.05M best
400K
AgenticMulti-step tasks with tools.
52.5
60.6 best
CodingCode writing and repair.
58.8
61.8 best
ReasoningLogic problems and puzzles.
69.8 best
67.0
MultimodalTasks with images and text.
69.9 best
57.3
KnowledgeFacts and expert knowledge.
65.0 best
64.1
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
86.6 best
59.5
MathMath problems.
60.4
—

Benchmarks

5 shared
Only Gemini 3.1 Pro Preview: 51Only GPT-5.3 Codex: 16
Gemini 3.1 Pro Preview
GPT-5.3 Codex
88.5% best6 / 142
87.3%16 / 142
32.0%80 / 123
61.8% best58 / 123
SWE-benchCoding
78.8% best28 / 87
78.0%33 / 87
67.4% best4 / 66
64.0%6 / 66
IOICoding
51.8%24 / 45
53.8% best22 / 45