BlogLLM Benchmark Scores as Published, Side by Side
GPT-5.5: Benchmark Scores as Published
Every result the catalog holds for GPT-5.5: 172 tests in 17 areas, from 14 pages. 107 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 2026-04-23Context window 1.05MMax output 128KInput Text, imageOutput TextWeights ClosedData version 2026103001, as of 2026-10-30
Knowledge Work
Agentic Coding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| SWE-bench Verified | 82.9% | MiniMax, MiniMax M3checked 2026-10-15 | Harness: Claude Code |
| SWE-Bench Pro | 58.6% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 7 more pages | Effort: xhigh |
| Terminal-Bench 2.1 | 85.6% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 7 more pages | - |
| DeepSWE v1.1 | 67% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 3 more pages | - |
| Terminal-Bench 2.0 | 82.7% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
| NL2Repo-Bench | 52.9 | MiniMax, MiniMax M3checked 2026-10-15; printed on 1 more page | - |
| SWE-Marathon | 14% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: xhigh; Dataset revision: H20-calibrated branch of the official tasks as of July 9, 2026, prior to the final v1.1 release |
| ProgramBench | 69.1% | Moonshot AI, Kimi K2.7 Code model cardchecked 2026-10-15; printed on 2 more pages | Effort: xhigh; Context: 262144 |
| SWE-Atlas-QnA (setting not stated) | 45.43% | MiniMax, MiniMax M3checked 2026-10-15 | - |
| FrontierSWE (dominance score) | 64.9 | Moonshot AI, Kimi K3 model cardchecked 2026-10-15; printed on 1 more page | Effort: xhigh |
| Kimi Code Bench 2.0 | 69% | Moonshot AI, Kimi K2.7 Code model cardchecked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 262144 |
| Artificial Analysis Coding Agent Index v1.1 | 76.4 | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| SWE-Atlas-Test Writing | 42.59% | MiniMax, MiniMax M3checked 2026-10-15 | - |
| DeepSWE v1.0 | 64.31% | xAI, Grok 4.5checked 2026-10-15 | Effort: xhigh; Harness: each provider's own harness (Artificial Analysis run) |
| SWE-fficiency | 46.6% | MiniMax, MiniMax M3checked 2026-10-15 | Harness: Claude Code |
| VIBE-V2 (MiniMax internal) | 50.5% | MiniMax, MiniMax M3checked 2026-10-15 | Harness: Claude Code |
| FrontierCode (Diamond) | 5.7% | Anthropic, Claude Fable 5 and Claude Mythos 5checked 2026-10-15 | Effort: xhigh |
| Expert-SWE (Internal) | 73.1% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
ML Engineering
Science and Math
Long Context
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AA-LCR | 74.3% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| MRCR v2 8-needle 128K | 94.8% | Google, Gemini 3.5 Flash model cardchecked 2026-10-15 | Context: 128k (average) |
| GraphWalks BFS 1M (F1) | 45.4% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 1M |
| GraphWalks BFS 256k (F1) | 73.7% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 256k |
| MRCR v2 8-needle 128K-256K | 87.5% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 128K-256K |
| MRCR v2 8-needle 512K-1M | 74% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 512K-1M |
| MRCR v2 8-needle 64K-128K | 83.1% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 64K-128K |
| MRCR v2 8-needle 256K-512K | 81.5% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 256K-512K |
| GraphWalks parents 1M (F1) | 58.5% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 1M |
| MRCR v2 8-needle 16K-32K | 96.5% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 16K-32K |
| MRCR v2 8-needle 32K-64K | 90% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 32K-64K |
| MRCR v2 8-needle 4K-8K | 98.1% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 4K-8K |
| MRCR v2 8-needle 8K-16K | 93% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 8K-16K |
| GraphWalks parents 256k (F1) | 90.1% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh; Context: 256k |
Computer Use
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| OSWorld-Verified | 78.7% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 5 more pages | Effort: xhigh |
| OSWorld 2.0 | 47.5% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 1 more page | - |
Multimodal Understanding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| CharXiv Reasoning (no tools) | 84.1% | Google, Gemini 3.5 Flash model cardchecked 2026-10-15; printed on 1 more page | Tools: none |
| MMMU-Pro | 81.2% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 4 more pages | Effort: xhigh; Tools: none |
| MathVision | 92.2% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: none |
| GDP.pdf | 26% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 1 more page | - |
| CharXiv Reasoning (with tools) | 89% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: enabled; Python |
| BabyVision (with tools) | 83.6% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: enabled; Python |
| MMMU-Pro (with tools) | 83.2% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 2 more pages | Effort: xhigh; Tools: with tools |
| OmniDocBench 1.5 | 87.5% | MiniMax, MiniMax M3checked 2026-10-15 | - |
| Video-MMMU | 86.4% | MiniMax, MiniMax M3checked 2026-10-15 | - |
| Video-MME (with subtitles) | 89.4% | MiniMax, MiniMax M3checked 2026-10-15; printed on 1 more page | - |
| MMVU | 81.7% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| MathVision (with tools) | 96.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: enabled; Python |
| ZeroBench (with tools, pass@5) | 41% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: enabled; Python |
| PerceptionBench | 55.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| ZeroBench (pass@5) | 22% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Tools: none |
| OmniDocBench (version not stated) | 89.4% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| WorldVQA ForceAnswer | 38.5% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
Cybersecurity
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| CyberGym | 81.8% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
| ExploitBench | 47.9% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15; printed on 1 more page | Harness: ExploitBench API harness, 5 seeds, reasoning continuity (footnote 2) |
| ExploitGym | 15.1% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| SEC-Bench Pro | 45.8% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| Production Chat | 0.928 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
| Synthetic Agentic Envs | 0.975 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
| Capture the Flag challenges | 88.1% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| Semi-Synthetic Agentic Envs | 0.963 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Humanity's Last Exam (no tools) | 41.4% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 5 more pages | Effort: xhigh; Tools: none |
| Humanity's Last Exam (with tools) | 52.2% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 4 more pages | Effort: xhigh; Tools: with tools |
| ARC-AGI-2 | 85% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| ARC-AGI-1 | 95% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
| ARC-AGI-3 | 0.43% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| CL-bench | 25.38% | MiniMax, MiniMax M3checked 2026-10-15 | - |
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| LiveSQLBench (Base-Full v1) | 40.17% | MiniMax, MiniMax M3checked 2026-10-15 | Harness: Claude Code |
| SVG-Bench (MiniMax internal) | 58.2% | MiniMax, MiniMax M3checked 2026-10-15 | Harness: Claude Code |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Toolathlon | 55.6% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 3 more pages | Effort: xhigh |
| τ³-bench Banking | 31.3% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| τ²-bench Telecom | 98% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
| Agents' Last Exam (pass rate) | 26.6% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh; Harness: Codex |
| Toolathlon Verified | 73.5% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
| MCP Atlas | 75.3% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 3 more pages | Effort: xhigh; Dataset revision: Scale AI, after the April 2026 update |
| MCP Atlas (public set) | 75.3% | MiniMax, MiniMax M3checked 2026-10-15; printed on 1 more page | - |
| Agents' Last Exam (score) | 46.9% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| BankerToolBench | 70.04% | MiniMax, MiniMax M3checked 2026-10-15 | - |
| MCPMark-Verified | 92.9% | Moonshot AI, Kimi K2.7 Code model cardchecked 2026-10-15; printed on 1 more page | Effort: xhigh; Context: 262144 |
| SaaS-Bench | 43.8% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
Web Search and Research
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BrowseComp | 84.4% | OpenAI, Introducing GPT-5.5checked 2026-10-15; printed on 3 more pages | Effort: xhigh |
| ResearchRubrics | 64% | Moonshot AI, Kimi K3 model cardchecked 2026-10-15 | Effort: xhigh |
Healthcare
Biosecurity
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Dual Use Safe | 0.926 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
| Severe Safe | 0.958 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
| Benign Not overrefuse | 0.917 | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19; printed on 1 more page | - |
| DNA sequence design for transcription factor binding (pass@1) | 13.82% | OpenAI, GPT-6 Astra System Cardchecked 2026-10-25 | Attempt budget: pass@1 |
| Hard-negative protein binding prediction (pass@4) | 1.5% | OpenAI, GPT-6 Astra System Cardchecked 2026-10-25 | Attempt budget: pass@4 |
Safety and Alignment
Engineering and Design
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BenchCAD | 44.4% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | - |
| Blueprint-Bench 2 (normalized score) | 36.2% | Anthropic, Claude Fable 5 and Claude Mythos 5checked 2026-10-15; printed on 1 more page | - |
| BenchCAD (python tool) | 55.8% | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 2026-10-15 | Tools: python tool |
Preference and Composite Rankings
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Artificial Analysis Intelligence Index v4.1 | 54.8 | OpenAI, GPT-5.6: Frontier intelligence that scales with your ambitionchecked 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. Cost per task is measured by Vals AI.
| Price | Value | Printed by | Setup as printed |
|---|---|---|---|
| Input | $5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $10 per 1M tokens above 272,000 tokens |
| Cached input | $0.5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $1 per 1M tokens above 272,000 tokens |
| Output | $30 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $45 per 1M tokens above 272,000 tokens |
| Input, batch | $2.5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $5 per 1M tokens above 272,000 tokens |
| Cached input, batch | $0.25 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $0.5 per 1M tokens above 272,000 tokens |
| Output, batch | $15 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $22.5 per 1M tokens above 272,000 tokens |
| Input, flex | $2.5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $5 per 1M tokens above 272,000 tokens |
| Cached input, flex | $0.25 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $0.5 per 1M tokens above 272,000 tokens |
| Output, flex | $15 | OpenAI, OpenAI API pricingchecked 2026-10-15 | Context: up to 272,000 tokens; Long context: $22.5 per 1M tokens above 272,000 tokens |
| Input, fast | $12.5 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cached input, fast | $1.25 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Output, fast | $75 | OpenAI, OpenAI API pricingchecked 2026-10-15 | - |
| Cost per taskVals AI, USD, lower is better | $16.66 | Vals AI, Vibe Code Bench v1.1 leaderboardchecked 2026-10-27 | Harness: OpenHands |
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 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
- 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.