Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

GPT-5.2 Codex

OpenAI

Strongest in Code generation (#6 of 77, on 1 of its 2 boards), weakest in Reasoning (#120 of 178). Above par in 3 of 11 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Claude Fable 5.1 and ahead of Qwen3.7 Max and Claude Sonnet 4.6.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Coding51.8#61/165
Code generation62.7#6/77
Agentic coding48#83/157
Core abilities48.4#114/204
Data analysis58.1#20/57
Instruction following47.4#34/57
Language39.5#47/57
General intelligence48.7#109/204
Reasoning45.9#120/178
Mathematics49.4#81/140
Reasoning41.6#105/140

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for GPT-5.2 Codex, left for the other.

Mistral Large 3won 7 · lost 0
Gemini 2.0 Flashwon 7 · lost 0
GPT-4o Miniwon 8 · lost 0
Qwen3.6 27Bwon 11 · lost 0
Grok 4.3won 11 · lost 0
GPT-5.4 Miniwon 11 · lost 0
Grok Build 0.1won 11 · lost 0
Gemini 3.5 Flash Litewon 10 · lost 1
MiniMax M3won 10 · lost 1
GPT-4.1 Nanowon 6 · lost 1
Ox Alphawon 9 · lost 2
Inklingwon 9 · lost 2
Kimi K2.7 Codewon 8 · lost 3
Qwen3.6 Pluswon 8 · lost 3
Granite 4.2 3Bwon 5 · lost 2
GPT OSS 120Bwon 5 · lost 2
K2 Horizon 7Bwon 5 · lost 2
GPT-4.1 Miniwon 5 · lost 2
GPT-5.4 Nanowon 7 · lost 4
Kimi K2.6won 7 · lost 4
GPT-5.6 Lunawon 7 · lost 4
GLM-5.3 Flashwon 7 · lost 4
DeepSeek V3won 5 · lost 3
Granite 4.2 8Bwon 4 · lost 3
K2 Horizon 3.7Bwon 4 · lost 3
Claude Haiku 4.5won 4 · lost 3
Nemotron 3 Ultrawon 4 · lost 3
GPT-4.1won 4 · lost 3
Claude Sonnet 4.6won 6 · lost 5
Qwen3.7 Maxwon 6 · lost 5
Gemini 3.6 Flashwon 5 · lost 6
Claude Opus 4.5won 5 · lost 6
Kimi K2won 3 · lost 4
Gemini 2.5 Flashwon 3 · lost 5
Gemini 2.5 Prowon 3 · lost 5
GPT-6 Lunawon 4 · lost 7
GLM-5.2won 4 · lost 7
Muse Spark 1.1won 4 · lost 7
Gemini 3.5 Flashwon 4 · lost 7
Claude Opus 4.6won 4 · lost 7
Qwen3.8 Flash Nextwon 4 · lost 7
Grok 4.5won 4 · lost 7
MiniMax M2.5won 2 · lost 5
Claude Sonnet 4.5won 2 · lost 5
Kimi K2 Thinkingwon 2 · lost 5
Grok 4 Fastwon 2 · lost 5
GPT-5 Miniwon 2 · lost 5
GLM-5won 2 · lost 5
Grok 4.20won 2 · lost 5
Kimi K2.5won 2 · lost 5
Gemini 3 Prowon 2 · lost 5
Claude Sonnet 5won 3 · lost 8
Union Alphawon 3 · lost 8
DeepSeek V4 Flashwon 3 · lost 8
GPT-5.2won 3 · lost 8
Qwen3.8 27Bwon 3 · lost 8
GLM-5.3won 3 · lost 8
Claude Opus 5won 3 · lost 8
Gemini 3.8 Flashwon 3 · lost 8
Gemini 3.1 Prowon 3 · lost 8
Claude Opus 4won 2 · lost 6
Claude Opus 4.8won 2 · lost 9
Muse Spark 1.2won 2 · lost 9
GPT-5.6 Terrawon 2 · lost 9
Gemini 3.7 Flashwon 2 · lost 9
Grok 4.6won 2 · lost 9
Grok 4.7won 2 · lost 9
DeepSeek V3.2won 1 · lost 6
Grok 3 Miniwon 1 · lost 6
Grok 4won 1 · lost 6
O3won 1 · lost 7
DeepSeek R1won 1 · lost 7
Claude Sonnet 4won 1 · lost 7
DeepSeek V4 Prowon 1 · lost 10
Claude Opus 4.7won 1 · lost 10
GPT-5.4won 1 · lost 10
DeepSeek V4.1 Flashwon 1 · lost 10
Qwen3.8 Maxwon 1 · lost 10
Kimi K3won 1 · lost 10
GPT-6 Solwon 1 · lost 10
Claude Opus 5.5won 1 · lost 10
Muse Spark 1.3won 1 · lost 10
GPT-5.5won 1 · lost 10
GPT-6 Astrawon 1 · lost 10
GPT-5won 0 · lost 7
O4 Miniwon 0 · lost 8
GPT-5.6 Solwon 0 · lost 11
Claude Fable 5.1won 0 · lost 11
Claude Fable 5won 0 · lost 11

§ 3 · Sources

Where the numbers come from

2 publications, 10 figures. Every one links to the page it was read from.

LiveBench ↗8 measuresread 2026-09-26
LiveBench 74LiveBench · Reasoning 77.7LiveBench · Coding 83.6LiveBench · Agentic Coding 49.4LiveBench · Mathematics 88.8LiveBench · Data Analysis 78.2LiveBench · Language 73.7LiveBench · Instruction Following 66.4
Vals.ai ↗2 measuresread 2026-09-26
Vals · SWE-bench Verified 72.4% (Mini-SWE-agent)Vals · Vibe Code Bench 37.91% (OpenHands)

Badge

PublicAI Index badge for GPT-5.2 Codex[![PublicAI Index](https://publicai.io/model-index/badge?model=gpt-5-2-codex)](https://publicai.io/model-index/m/gpt-5-2-codex)