Gemini 4 Argon
Claude Opus 5.5
Muse Wonder
MiMo-V2.6-Pro
Qwen3.8-Max
Kimi K3
GLM-5.3
ERNIE 5.1
DeepSeek-V4.1-Flash-20260910
Grok 4.20 Beta
amazon-nova-experimental-chat
Dola Seed 2.0 Pro
NVIDIA-Nemotron-3-Ultra-550B-A55B
Gemma 4 31B
Step 5 Preview
GPT-5.2
GPT-5.2 is a proprietary model from OpenAI, released Dec 10, 2025. It ranks #130 of 351 on the LMArena text leaderboard with a rating of 1416. It costs $1.75 per million input tokens and $14 per million output tokens, with a 400K-token context window.
- Developer
- OpenAI
- Released
- Dec 10, 2025
- Context window
- 400K
- Price in / out, per 1M tokens
- $1.75 / $14
- Weights
- Proprietary
- Inputs
- Files, Images, Text
- Reasoning
- Yes
- Tool use
- Yes
The rating LMArena published at each update since Dec 18, 2025. Ratings move as votes come in and new models join.
Benchmarks
| Benchmark | Score bar | Accuracy | Rank | Cost / test |
|---|---|---|---|---|
| SWE-benchCoding | 75.80% | 40 / 85 | $1.45 | |
| Vibe Code BenchCoding | 53.50% | 56 / 115 | $17.75 | |
| LiveCodeBenchCoding | 85.36% | 33 / 143 | $1.75 / $14 | |
| MedCodeMedicine | 49.75% | 20 / 107 | $1.75 / $14 | |
| MedQAMedicine | 94.13% | 19 / 92 | $1.75 / $14 | |
| GPQA DiamondReasoning | 91.67% | 20 / 138 | $1.75 / $14 | |
| MMLU-ProReasoning | 86.23% | 46 / 138 | $1.75 / $14 | |
| AIMEReasoning | 96.88% | 2 / 96 | $1.75 / $14 | |
| MMMUReasoning | 86.67% | 17 / 93 | $1.75 / $14 |
Vals AI, as of Oct 24, 2026.
LMArena, by category
| Leaderboard | Rating | Rank | Votes |
|---|---|---|---|
| TextLMArena | 1416 | 130 / 351 | 49.6K |
| CodingLMArena | 1442 | 116 / 346 | 12.2K |
| MathLMArena | 1440 | 86 / 339 | 3.1K |
| Creative writingLMArena | 1376 | 136 / 349 | 7.7K |
| Instruction followingLMArena | 1409 | 107 / 351 | 15.3K |
| Hard promptsLMArena | 1428 | 114 / 351 | 29.9K |
| Multi-turnLMArena | 1419 | 116 / 349 | 9.2K |
| Long queriesLMArena | 1413 | 124 / 330 | 34.1K |
| ExpertLMArena | 1445 | 90 / 305 | 4K |
| Legal and governmentIndustries | 1427 | 117 / 325 | 3.8K |
| Medicine and healthcareIndustries | 1425 | 126 / 325 | 6.2K |
| Business and financeIndustries | 1414 | 116 / 344 | 9.8K |
| ChineseLanguages | 1447 | 131 / 328 | 3.6K |
| SpanishLanguages | 1414 | 127 / 249 | 1.5K |
| GermanLanguages | 1444 | 63 / 261 | 834 |
| FrenchLanguages | 1454 | 77 / 252 | 1.4K |
| JapaneseLanguages | 1420 | 53 / 236 | 446 |
| WebDevLMArena | 1417 | 91 / 124 | 1.1K |
| VisionLMArena | 1246 | 68 / 146 | 19K |
LMArena, CC BY 4.0, as of Oct 22, 2026. Where a model is rated at several effort levels, its best is shown (GPT-5.2 (High), GPT-5.2).
Compare GPT-5.2
OpenAI
The lab behind GPT and ChatGPT, which brought large language models to hundreds of millions of people. Founded 2015, based in San Francisco. openai.com
Questions
Who makes GPT-5.2?
GPT-5.2 is made by OpenAI, founded 2015 and based in San Francisco.
How much does GPT-5.2 cost?
$1.75 per million input tokens and $14 per million output tokens, at its provider’s list price.
What is GPT-5.2’s context window?
400K tokens (400,000).
How does GPT-5.2 rank?
#130 of 351 on LMArena’s text leaderboard, rated 1416 (95% interval 1413–1420) from 49,643 votes, as of Oct 22, 2026.
Is GPT-5.2 open source?
No. GPT-5.2 is proprietary and served through its provider’s API.