GPT-5.2
GPT-5.2 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family.
text + image + file → text · made by OpenAI
$1.75→$14per Mtok in / out
OpenAI · 10 sellers
Sold by 21 ways
| Seller | Lane | Rate |
|---|---|---|
| Microsoft Azure AI cloud | batch | $0.88per Mtok in$7per Mtok out$0.087per Mtok cachedprices.azure.com · read 2026-08-24 |
| Databricks cloud | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cached$1.75per Mtok cache writemodels.dev · read 2026-09-16 |
| DigitalOcean cloud | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedmodels.dev · read 2026-09-16 |
| Microsoft Azure AI cloud | standard | $1.75per Mtok in$14per Mtok out$0.12per Mtok cachedmodels.dev · read 2026-09-16 |
| OpenAI api | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cacheddevelopers.openai.com · read 2026-08-24 |
| Microsoft Azure AI cloud | priority | $3.5per Mtok in$28per Mtok out$0.35per Mtok cachedraw.githubusercontent.com · read 2026-09-16 |
| Requesty 3 ways | $0.79per Mtok in$6.3per Mtok out$0.079per Mtok cached | |
| Requesty aggregatornot on the seller's list today | openai flex | $0.79per Mtok in$6.3per Mtok out$0.079per Mtok cachedrouter.requesty.ai · read 2026-08-25 |
| Requesty aggregatornot on the seller's list today | openai-responses flex | $0.79per Mtok in$6.3per Mtok out$0.079per Mtok cachedrouter.requesty.ai · read 2026-08-25 |
| Requesty aggregator | openai | $0.88per Mtok in$7per Mtok out$0.087per Mtok cachedrouter.requesty.ai · read 2026-09-16 |
| Requesty aggregator | azure | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedrouter.requesty.ai · read 2026-08-28 |
| Requesty aggregator | openai-responses | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedrouter.requesty.ai · read 2026-09-10 |
| Nous Research 2 ways | $1.75per Mtok in$14per Mtok out$0.17per Mtok cached | |
| Nous Research aggregator | batch | $0.88per Mtok in$7per Mtok out$0.087per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| Nous Research aggregator | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| OpenRouter 4 ways | $1.75per Mtok in$14per Mtok out$0.17per Mtok cached | |
| OpenRouter aggregator | batch | $0.88per Mtok in$7per Mtok out$0.087per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregator | openai/flex | $0.88per Mtok in$7per Mtok out$0.087per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregator | openai/fast | $3.5per Mtok in$28per Mtok out$0.35per Mtok cachedopenrouter.ai · read 2026-08-25 |
| ElectronHub aggregator | standard | $1.25per Mtok in$10per Mtok outapi.electronhub.ai · read 2026-09-16 |
| GitHub Copilot 0 ways | $1.75per Mtok in$14per Mtok out$0.17per Mtok cached | |
| GitHub Copilot aggregatornot on the seller's list today | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedmodels.dev · read 2026-09-04 |
| Vercel AI Gateway aggregator | standard | $1.75per Mtok in$14per Mtok out$0.17per Mtok cachedai-gateway.vercel.sh · read 2026-08-25 |
| Venice AI aggregator | standard | $2.19per Mtok in$17.5per Mtok out$0.22per Mtok cachedmodels.dev · read 2026-09-16 |
Measured 62 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 1stof 61 | AIME 2025 | Accuracy (± 95% CI) (xhigh) | 100 |
| 1stof 18 | Algotune — Epoch AI | Score (medium) | 2.05 |
| 1stof 11 | Gdpval — Epoch AI | Win Rate (%) (none) | 0.497 |
| 2ndof 38 | Vpct — Epoch AI | Correct (xhigh) | 0.84 |
| 3rdof 8 | τ²-bench Leaderboard (Overall: Retail · Airline · Telecom) | Pass^1 | 84.8 |
| 4thof 32 | AIME 2026 | Accuracy (± 95% CI) (high) | 98.33 |
| 4thof 50 | Metr time horizons — Epoch AI | average_score (high) | 0.753 |
| 8thof 13 | Rli — Epoch AI | Score (medium) | 2.502 |
| 9thof 101 | FrontierMath (Tiers 1-3) | mean_score (xhigh) | 40.7 |
| 9thof 47 | SWE-bench Verified (default "Bash Only" view, agent = mini-SWE-agent) | % Resolved | 72.8 |
| 9thof 47 | SWE-bench Verified (default "Bash Only" view, agent = mini-SWE-agent) | % Resolved (high) | 72.8 |
| 10thof 21 | Ebr bench — Epoch AI | mean_score (xhigh) | 23.016 |
| 10thof 38 | Gso — Epoch AI | Score OPT@1 (high) | 0.275 |
| 10thof 51 | Humanity's Last Exam (Epoch AI replication) | Accuracy | 27.8 |
| 10thof 262 | MMLU-Pro | Overall (accuracy) | 87.4 |
| 11thof 34 | LMArena · Search | Score (Elo) | 1207 |
| 12thof 222 | Chess puzzles — Epoch AI | mean_score (xhigh) | 49 |
| 12thof 23 | Cl bench — Epoch AI | Overall | 0.182 |
| 13thof 72 | Frontiermath tier 4 — Epoch AI | mean_score (xhigh) | 18.8 |
| 13thof 28 | τ³-Banking Leaderboard | Pass^1 | 32.2 |
| 15thof 46 | Enigma eval — Epoch AI | Accuracy | 10.39 |
| 16thof 109 | Berkeley Function-Calling Leaderboard (BFCL) V4 | Overall Acc (fc) | 55.87 |
| 16thof 25 | SWE-Bench Pro (Public Dataset) | Resolve Rate | 29.94 |
| 19thof 65 | Apex agents — Epoch AI | Pass@1 score (xhigh) | 0.344 |
| 19thof 35 | Swe bench verified — Epoch AI | mean_score (high) | 73.76 |
| 20thof 106 | Frontiermath tiers 1 3 v2 — Epoch AI | mean_score (xhigh) | 67.4 |
| 25thof 112 | Ale bench — Epoch AI | Performance (high) | 1293.55 |
| 25thof 161 | Dtbench — Epoch AI | Accuracy (xhigh) | 90.93 |
| 27thof 313 | GPQA Diamond | mean_score (xhigh) | 91.4 |
| 28thof 42 | Design Arena · godotgamedev (agents) | Elo | 1142 |
| 28thof 62 | Frontiermath tier 4 v2 — Epoch AI | mean_score (xhigh) | 31.7 |
| 28thof 123 | Lmca — Epoch AI | Score (xhigh) | 43.91 |
| 28thof 290 | OTIS Mock AIME 2024-2025 | mean_score (high) | 96.111 |
| 28thof 170 | Weirdml — Epoch AI | Accuracy (xhigh) | 72.19 |
| 30thof 81 | Forecastbench — Epoch AI | Overall score | 60.1 |
| 34thof 202 | Terminalbench — Epoch AI | Accuracy mean (medium) | 0.649 |
| 36thof 41 | Deepresearchbench — Epoch AI | Average score (low) | 0.411 |
| 38thof 60 | Vending bench 2 — Epoch AI | Score | 3591.326 |
| 41stof 127 | Mystery game puzzles — Epoch AI | mean_score (high) | 23 |
| 42ndof 80 | Simpleqa verified — Epoch AI | mean_score (xhigh) | 37.1 |
| 42ndof 105 | Vectara Hallucination Leaderboard | Hallucination Rate | 8.4 |
| 45thof 55 | Design Arena · androidnative (agents) | Elo | 1074 |
| 46thof 64 | Proofbench — Epoch AI | Accuracy (xhigh) | 15 |
| 47thof 121 | Webdev arena — Epoch AI | Arena Score (high) | 1480 |
| 50thof 152 | LMArena · Vision | Score (Elo) (high) | 1244 |
| 52ndof 92 | Design Arena · asciiart (models) | Elo | 1170 |
| 53rdof 63 | Design Arena · mobileapps (agents) | Elo | 1121 |
| 53rdof 62 | Design Arena · webapps (agents) | Elo | 1100 |
| 55thof 64 | Design Arena · fullstack (agents) | Elo | 1052 |
| 62ndof 102 | Simplebench — Epoch AI | Score (AVG@5) (high) | 0.458 |
| 64thof 155 | Design Arena · gamedev (models) | Elo | 1218 |
| 66thof 109 | Design Arena · svg (models) | Elo | 1164 |
| 69thof 152 | Design Arena · dataviz (models) | Elo | 1217 |
| 70thof 128 | LMArena · WebDev | Score (Elo) | 1417 |
| 73rdof 150 | Design Arena · uicomponent (models) | Elo | 1205 |
| 75thof 162 | Design Arena · website (models) | Elo | 1207 |
| 78thof 227 | Arc agi 2 — Epoch AI | Score (xhigh) | 52.91 |
| 81stof 244 | Arc agi — Epoch AI | Score (xhigh) | 86.2 |
| 88thof 156 | Design Arena · codecategories (models) | Elo | 1185 |
| 89thof 402 | LMArena · Text | Score (Elo) (high) | 1438 |
| 106thof 147 | Design Arena · 3d (models) | Elo | 1107 |
| 167thof 656 | Epoch capabilities index — Epoch AI | ECI Score (medium) | 153.43 |
About
GPT-5.2 — a text model from OpenAI, sold by 10 companies from $1.25 in and $10 out per million tokens, placed 1st of 61 on AIME 2025.
It takes text, images and files and returns text, with a context window of 400,000 tokens. It was published in December 2025, trained on material up to August 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning, code and agents. Ten companies sell it. The cheapest is $1.25 in and $10 out per million tokens at ElectronHub. Beside the standard rate there are batch, faster, flex and separately routed lanes. It has been measured on 61 boards, and stands best at 1st of 61 on AIME 2025.
Every current figure
- Maker
- OpenAI
- Register
- model
- Takes
- text + image + file
- Returns
- text
- Context
- 400,000 tokens
- Longest answer
- 128,000 tokens
- Published
- December 2025
- Knowledge to
- August 2025
- Licence
- not read
- Sellers
- 10
- Price
- $1.25 in and $10 out per million tokens — ElectronHub
- Boards
- 61
- Best place
- 1st of 61 — AIME 2025
Known as 18 names
GPT-5.2 (high)GPT-5.2 (xhigh)GPT-5.2 (Refine.)GPT-5.2-2025-12-11 (FC)gpt-5.2-searchgpt-5.2-search-non-reasoningGPT 5.2GPT 5.2 (2025-12-11) (high)GPT 5.2 (high)GPT 5.2 high5.2openai/gpt-5.2openai/gpt-5.2:batchazure/gpt-5.2@eastus2openai-responses/gpt-5.2openai-responses/gpt-5.2:flexopenai/gpt-5.2:flexOpenAI: GPT-5.2 (batch)