BlogLLM Benchmark Scores as Published, Side by Side
GPT-5.2: Benchmark Scores as Published
Every result the catalog holds for GPT-5.2: 126 tests in 16 areas, from 12 pages. 50 printed by OpenAI itself, the rest by the company named beside each number. Each number is as printed; nothing is ranked or rescored.
Developer OpenAIReleased 2025-12-11Context window 400KMax output 128KInput Text, imageOutput TextWeights ClosedData version 2026103001, as of 2026-10-30
Knowledge Work
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| APEX-Agents | 23% | Google, Gemini 3.1 Pro model cardchecked 2026-10-15 | Effort: xhigh |
| GDPval-AA (Elo) | 1462 | Anthropic, Claude Opus 4.6checked 2026-10-15; printed on 2 more pages | - |
| Vals Finance Agent v1.1 | 59.5% | OpenAI, Introducing GPT-5.4checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| Vending-Bench 2 (final balance in USD) | $3591.33 | Z.ai, GLM-5 model cardchecked 2026-10-15 | Effort: xhigh |
| OfficeQA | 63.1% | OpenAI, Introducing GPT-5.4checked 2026-10-15 | Effort: xhigh |
| GDPval (wins or ties) | 70.9% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 2 more pages | Effort: heavy |
| Investment Banking Modeling Tasks (Internal) | 68.4% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: heavy |
| GDPval (no ties) | 61% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: heavy |
| GDPval (ties allowed, clear wins) | 49.8% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: heavy |
Agentic Coding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| SWE-bench Verified | 80% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 6 more pages | Effort: xhigh |
| SWE-Bench Pro | 55.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 5 more pages | Effort: xhigh |
| SWE-bench Multilingual | 72% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 2 more pages | - |
| Terminal-Bench 2.0 | 62.2% | OpenAI, Introducing GPT-5.2-Codexchecked 2026-10-15; printed on 8 more pages | - |
| SWE-Lancer IC Diamond | 74.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Dataset revision: 197 of 237 IC Diamond problems (40 omitted) |
ML Engineering
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| PaperBench | 63.7 | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Effort: xhigh |
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPQA Diamond | 92.4% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 6 more pages | Effort: xhigh; Tools: none |
| AIME 2025 | 100% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 2 more pages | Effort: xhigh; Tools: none |
| SciCode | 52% | Google, Gemini 3.1 Pro model cardchecked 2026-10-15; printed on 2 more pages | Effort: xhigh |
| HMMT February 2025 | 99.4% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 2 more pages | Effort: xhigh; Tools: none |
| IMO-AnswerBench | 86.3% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 2 more pages | - |
| AIME 2026 | 96.7% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| HMMT November 2025 | 100% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | - |
| FrontierMath Tier 1-3 | 40.3% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Tools: python |
| PolyMATH | 62.5% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| FrontierMath Tier 4 | 14.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Tools: python |
| Frontier Science Research | 25.2% | OpenAI, Introducing GPT-5.4checked 2026-10-15 | Effort: xhigh |
Long Context
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AA-LCR | 72.7% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | - |
| LongBench v2 | 54.5% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | - |
| MRCR v2 8-needle 128K | 83.8% | Google, Gemini 3.1 Pro model cardchecked 2026-10-15 | Effort: xhigh; Context: 128k (average) |
| GraphWalks BFS up to 128k (F1) | 94% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: <128k |
| MRCR v2 8-needle 128K-256K | 77% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 128k-256k |
| MRCR v2 8-needle 64K-128K | 85.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 64k-128k |
| GraphWalks parents up to 128k (accuracy) | 89% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: <128k |
| MRCR v2 8-needle 16K-32K | 95.3% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 16k-32k |
| MRCR v2 8-needle 32K-64K | 92% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 32k-64k |
| MRCR v2 8-needle 4K-8K | 98.2% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 4k-8k |
| MRCR v2 8-needle 8K-16K | 89.3% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 8k-16k |
| BrowseComp Long Context 128k | 92% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: xhigh; Context: 128k |
| BrowseComp Long Context 256k | 89.8% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: xhigh; Context: 256k |
Computer Use
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| OSWorld-Verified | 37.9% | OpenAI, Introducing GPT-5.3-Codexchecked 2026-10-15; printed on 3 more pages | Effort: xhigh |
| ScreenSpot-Pro (with tools) | 86.3% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: xhigh; Tools: python |
Multimodal Understanding
Cybersecurity
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Capture the Flag challenges | 67.7% | OpenAI, Introducing GPT-5.3-Codexchecked 2026-10-15 | Effort: xhigh |
| SecCodeBench | 68.7% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
General Knowledge and Reasoning
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| LiveCodeBench v6 | 87.7% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| LiveCodeBench Pro (Elo) | 2393 | Google, Gemini 3.1 Pro model cardchecked 2026-10-15 | Effort: xhigh |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Toolathlon | 46.3% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 3 more pages | Effort: xhigh |
| τ²-bench Telecom | 98.7% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 4 more pages | Effort: xhigh |
| MCP Atlas | 60.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 4 more pages | Effort: xhigh |
| τ²-bench Retail | 82% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 3 more pages | Effort: xhigh |
| MCP Atlas (public set) | 68% | Z.ai, GLM-5 model cardchecked 2026-10-15 | Effort: xhigh |
| MCPMark | 57.5% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| τ²-bench (domain not stated) | 87.1% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | - |
| DeepPlanning | 44.6% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| VitaBench | 38.2% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| BFCL v4 | 63.1% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
Web Search and Research
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BrowseComp | 65.8% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 5 more pages | Effort: xhigh |
| BrowseComp-ZH | 76.1% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | Tools: enabled; search; Context: 256000 |
| WideSearch (metric not stated) | 76.8 | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | Tools: enabled; search; Context: 256000 |
| Seal-0 | 45% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | Tools: enabled; search; Context: 256000 |
| DeepSearchQA (accuracy) | 71.3% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15 | Effort: xhigh; Tools: search, code interpreter, browsing |
Instruction Following
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| IFBench | 75.4 | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 1 more page | - |
| MultiChallenge | 57.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| IFEval | 94.8% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
Multilingual
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MMMLU | 89.6% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 4 more pages | Effort: xhigh |
| MMLU-ProX | 83.7% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| WMT24++ (XCOMET-XXL) | 78.8 | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Global PIQA | 90.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| INCLUDE | 87.5% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| MaXIFE | 88.4% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| NOVA-63 | 54.6% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
Healthcare
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MedXpertQA MM | 73.3% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| PMC-VQA | 58.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| SLAKE | 76.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
Safety and Alignment
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Sensitive conversations (emotional reliance) | 0.955 | OpenAI, Introducing GPT-5.2checked 2026-10-15 | - |
| Sensitive conversations (mental health) | 0.915 | OpenAI, Introducing GPT-5.2checked 2026-10-15 | - |
| Sensitive conversations (self-harm) | 0.963 | OpenAI, Introducing GPT-5.2checked 2026-10-15 | - |
Price
API list prices in US dollars per one million tokens, as printed on the developer's pricing page on the day checked.
| Price | Value | Printed by | Setup as printed |
|---|---|---|---|
| Input | $1.75 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cached input | $0.175 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Output | $14 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Input, batch | $0.875 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cached input, batch | $0.0875 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Output, batch | $7 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Input, flex | $0.875 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cached input, flex | $0.0875 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Output, flex | $7 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Input, fast | $3.5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cached input, fast | $0.35 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Output, fast | $28 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
More from OpenAI
- GPT-6.1 Terra
- GPT-6.1 Sol
- GPT-6 Luna
- GPT-6 Sol
- GPT-6 Astra
- GPT-5.6 Luna
- GPT-5.6 Sol
- GPT-5.6 Terra
- GPT-5.5
- GPT-5.5 Pro
- GPT-5.4 mini
- GPT-5.4 nano
- GPT-5.4
- GPT-5.4 Pro
- GPT-5.3-Codex
- GPT-5.2-Codex
- GPT-5.2 Instant
- GPT-5.2 Pro
- GPT-5.1-Codex-Max
- GPT-5.1
- GPT-5.1 Instant
- GPT-5-Codex
- GPT-5
- GPT-5 mini
- GPT-5 Pro
- gpt-oss-120b
- gpt-oss-20b
- ChatGPT agent
- codex-1
- o3
- OpenAI o4-mini
- GPT-4.1
- GPT-4.5
- Deep research
- o3-mini
- o1
- o1-mini
- GPT-4o
- GPT-5.5 Instant
- gpt-5.6-cyber
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.