Claude Opus 5 vs GPT-5.6 Terra
Full profiles: Claude Opus 5 (Anthropic) and GPT-5.6 Terra (OpenAI): every benchmark, price and rank.
Updated · Data from 11 sources
As of 25 Oct 2026, Claude Opus 5 (max) leads on quality (68.8 vs 64.4 on the BenchLeader Index), but GPT-5.6 Terra (max) costs 2.2× less, so it is the better buy unless you need the extra quality. Pick Claude Opus 5 for knowledge (72 vs 57); pick GPT-5.6 Terra for long context (67 vs 65).
Claude Opus 5 benchmarks max Anthropic
GPT-5.6 Terra benchmarks max OpenAIOpen in the comparison toolWhich to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | Claude Opus 5 | 70 vs 65 category score; SciCode 56.4% vs 55.0% | Medium |
| Agents & tool use | Claude Opus 5 | 64 vs 59 category score; Terminal-Bench 51.8% vs 21.5% | Medium |
| Maths | Claude Opus 5 | 71 vs 69 category score; FrontierMath Tier 4 73.2% vs 70.7% | Medium |
| Reasoning | Claude Opus 5 | 74 vs 70 category score; GPQA Diamond 93.9% vs 93.3% | Medium |
| Knowledge | Claude Opus 5 | 72 vs 57 category score; SimpleQA Verified 59.9% vs 43.2% | High |
| Long documents | GPT-5.6 Terra | 67 vs 65 category score; AA-LCR 83.0% vs 79.3% | Medium |
| Images and other media | Claude Opus 5 | 68 vs 64 category score; MMMU-Pro 84.7% vs 80.7% | Medium |
| Chat quality (human preference) | – | Only Claude Opus 5 has an Arena rating (69 category score) | Not comparable |
| Lowest cost | GPT-5.6 Terra | $4.50 vs $10.00 per 1M blended; agentic coding session $0.168 vs $0.400 | – |
| Fastest first answer | Claude Opus 5 | 24.1 s (Claude Opus 5 (max)) vs 101.0 s (GPT-5.6 Terra (max)) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Claude Opus 5 (max) leads on quality: 68.8 vs 64.4.
- Claude Opus 5 (max) is stronger in agents & tools, coding, composite, human preference, knowledge, maths, multimodal, reasoning.
- GPT-5.6 Terra (max) is stronger in long context, instruction following.
- GPT-5.6 Terra (max) is 2.2× cheaper ($4.50 vs $10.00 per 1M blended).
- GPT-5.6 Terra (max) streams 2.2× faster (126 vs 57 tokens per second).
At a glance
| Spec | Claude Opus 5 | GPT-5.6 Terra |
|---|---|---|
| Developer | Anthropic | OpenAI |
| Released | 24 Jul 2026 | 9 Jul 2026 |
| Context window (tokens) | 1,000,000 | 1,050,000 |
| Compared at | Claude Opus 5 (max) | GPT-5.6 Terra (max) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $5.00 / $25.00 / $0.500 | $2.00 / $12.00 / $0.200 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Claude Opus 5 (max) | GPT-5.6 Terra (max) | Leader |
|---|---|---|---|
| Reasoning | 74.1 | 70.1 | Claude Opus 5 |
| Coding | 70.2 | 65.1 | Claude Opus 5 |
| Agents & tools | 64.3 | 58.9 | Claude Opus 5 |
| Maths | 70.6 | 68.9 | Claude Opus 5 |
| Knowledge | 71.6 | 57.2 | Claude Opus 5 |
| Instruction following | – | 69.9 | not comparable |
| Human preference | 68.5 | – | not comparable |
| Multimodal | 67.9 | 63.7 | Claude Opus 5 |
| Long context | 64.7 | 66.6 | GPT-5.6 Terra |
| Composite | 86.3 | 76.6 | Claude Opus 5 |
Where sources disagree
Across the 41 benchmarks both have, Claude Opus 5 leads 31, GPT-5.6 Terra leads 10. Every overall index that rates both puts Claude Opus 5 ahead.
| Overall index | Claude Opus 5 (max) | GPT-5.6 Terra (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 68.8 | 64.4 | Claude Opus 5 |
| AA Intelligence Index v4.3.2 | 50.8 | 42.1 | Claude Opus 5 |
| Arena Text | 1491 | 1467† | Claude Opus 5 |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.
| Measure | Claude Opus 5 (max) | GPT-5.6 Terra (max) |
|---|---|---|
| Blended price per 1M tokens | $10.00 | $4.50 |
| Input / output per 1M | $5.00 / $25.00 | $2.00 / $12.00 |
| Cached input per 1M | $0.500 | $0.200 |
| Price per task: chat reply | $0.0095 | $0.0044 |
| Price per task: summarize a 30-page report | $0.075 | $0.031 |
| Price per task: code edit | $0.068 | $0.030 |
| Price per task: agentic coding session | $0.400 | $0.168 |
| Price per task: structured extraction | $0.015 | $0.0064 |
| Cost to run a full benchmark suite | $5.86 | $1.40 |
| Output speed | 57 tok/s | 126 tok/s |
| Time to first answer | 24.1 s | 101.0 s |
| Time to a complete answer | 32.9 s | 105.0 s |
Per token is not per task: on Artificial Analysis' full suite Claude Opus 5 (max) cost $5.86 and GPT-5.6 Terra (max) $1.40, because Claude Opus 5 (max) used about 1.9× as many tokens for the same work, which a per-token price hides.
Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 5 (max) scores 68.8 and takes 24 s to its first answer token; Claude Opus 5 (high) scores 67.9 in 10 s, Claude Opus 5 (xhigh) scores 67.4 in 25 s. GPT-5.6 Terra (max) scores 64.4 and takes 101 s to its first answer token; GPT-5.6 Terra (xhigh) scores 62.8 in 7 s, GPT-5.6 Terra (high) scores 61.3 in 2 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
137 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Claude Opus 5 (max) | GPT-5.6 Terra (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 68.8 | 64.4 |
| Agents & tools score | BenchLeader | 64.3 | 58.9 |
| Coding score | BenchLeader | 70.2 | 65.1 |
| Composite score | BenchLeader | 86.3 | 76.6 |
| Human preference score | BenchLeader | 68.5 | – |
| Knowledge score | BenchLeader | 71.6 | 57.2 |
| Long context score | BenchLeader | 64.7 | 66.6 |
| Maths score | BenchLeader | 70.6 | 68.9 |
| Multimodal score | BenchLeader | 67.9 | 63.7 |
| Reasoning score | BenchLeader | 74.1 | 70.1 |
| Instruction following score | BenchLeader | – | 69.9 |
| Blended price $/M | BenchLeader | $10.00 | $4.50 |
| Output speed | BenchLeader | 57 tok/s | 126 tok/s |
| Time to first answer | BenchLeader | 24.1 s | 101.0 s |
| Context window | BenchLeader | 1,000,000 tokens | 1,050,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 93.9% | 93.3% |
| SimpleBench | SimpleBench | 80.6%† | 48.9%† |
| Arena Hard Prompts | Arena (LMArena) | 1515 | 1490† |
| LiveBench 2.0 Reasoning | LiveBench | 91.2%† | 90.6%† |
| LiveBench 1.0 Reasoning | LiveBench | 91.2% | 90.6% |
| GPQA Diamond (AA) | Artificial Analysis | 93.2% | 92.5% |
| Humanity's Last Exam (AA) | Artificial Analysis | 54.9% | 42.9% |
| GPQA Diamond (Vals) | Vals AI | 93.4%† | 90.9%† |
| Kagi LLM Benchmark | Kagi LLM Benchmark | – | 51.3%† |
| ARC-AGI-1 | ARC Prize | 97.5% | 96.5% |
| ARC-AGI-2 | ARC Prize | 90.4% | 83.9% |
| ARC-AGI-3 | ARC Prize | 30.2%† | 0.8% |
| CritPt | Artificial Analysis | 29.1% | 30.0% |
| MysteryMechanism | Vals AI | 37.4%† | – |
| Chess Puzzles | Epoch AI Benchmarking Hub | 43.0% | 47.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 59.0% | 35.0% |
| LMCA | Epoch AI Benchmarking Hub | 63.3% | 55.0% |
| DTBench | Epoch AI Benchmarking Hub | 97.6% | 93.3% |
| ForecastBench | ForecastBench | 66.5% | – |
| BTF-3 | Epoch AI Benchmarking Hub | 12.0%† | – |
| Coding | |||
| SciCode | SciCode | 56.4% | 55.0% |
| WeirdML | WeirdML | 91.8% | 78.3%† |
| FrontierCode | Cognition | 53.4% | 41.3%† |
| Arena Coding | Arena (LMArena) | 1528 | 1518† |
| Arena WebDev | Arena (LMArena) | 1691 | 1521† |
Show all 137 benchmarks
| Metric | Publisher | Claude Opus 5 (max) | GPT-5.6 Terra (max) |
|---|---|---|---|
| Coding | |||
| LiveBench 2.0 Coding | LiveBench | 90.8%† | 86.2%† |
| LiveBench 1.0 Coding | LiveBench | 81.5% | 78.3% |
| LiveCodeBench | Vals AI | 89.0%† | 85.9%† |
| IOI | Vals AI | 84.3%† | 87.6% |
| SWE-bench (Vals) | Vals AI | 97.0%† | 95.4% |
| Code Migration | Vals AI | 57.5%† | 47.8% |
| ProgramBench | Vals AI | 3.0%† | 0.5% |
| Vibe Code Bench 1-100 | Vals AI | 28.5%† | 14.8% |
| Vibe Code Bench v1.1 | Vals AI | 88.4%† | 74.6% |
| SWE Atlas: Codebase QnA | Scale AI SEAL | 63.2%† | – |
| SWE Atlas: Test Writing | Scale AI SEAL | 62.2%† | – |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | – | 83.8%† |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 73.7% | 69.6% |
| CursorBench | Cursor | 46.6% | 41.3% |
| ALE-Bench | Epoch AI Benchmarking Hub | 2164.6† | 1951.4 |
| FrontierSWE | Epoch AI Benchmarking Hub | 52.0% | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 49.0% | 35.4% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 89.1% | 88.0% |
| Startrise Frontend | Startrise | 82.3%† | 66.9%† |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 51.8% | 21.5% |
| OSWorld-Verified 2.0 | OSWorld | 31.4% | – |
| APEX-Agents | Mercor | 65.8% | 58.2% |
| Arena Agent | Arena (LMArena) | 8.2 | 0.2† |
| LiveBench Long Agentic | LiveBench | 62.1%† | 48.4%† |
| LiveBench Browser | LiveBench | 84.0%† | 84.0%† |
| LiveBench Agentic Coding | LiveBench | 65.2% | 55.0% |
| Terminal-Bench Hard | Artificial Analysis | – | 57.6% |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 86.3% |
| Terminal-Bench 2.1 (Vals) | Vals AI | 84.6%† | 77.5% |
| MCP Atlas | Scale AI SEAL | 85.8%† | – |
| HiL-Bench | Scale AI SEAL | 57.0%† | – |
| τ³-bench Banking | τ²-bench | 48.7% | – |
| GDPval-AA v2.1 | Artificial Analysis | – | 47.6% |
| τ³-Banking (AA) | Artificial Analysis | 42.1% | 40.2% |
| ITBench SRE (AA) | Artificial Analysis | – | 51.0% |
| Analyst Agent (AA) | Artificial Analysis | 53.8% | – |
| APEX-Agents (AA) | Artificial Analysis | – | 38.9% |
| CUA-bench | Vals AI | 9.0% | – |
| CyberBench | Vals AI | 65.4%† | 72.1% |
| Finance Agent v2 | Vals AI | 58.6%† | 54.4% |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 6.7%† | 0.8% |
| Legal Research Bench | Vals AI | 55.3%† | 41.4% |
| SkillsBench | Vals AI | 60.4%† | 58.9% |
| SREBench | Vals AI | 12.2%† | 9.5% |
| Tax Agent Bench | Vals AI | 45.5%† | 30.0% |
| Terminal-Bench 4.0 (Vals) | Vals AI | 53.5%† | 22.7% |
| Terminal-Bench Science | Vals AI | 27.1%† | 10.0% |
| Time Horizon Index: KSP | Vals AI | 18.8%† | – |
| Arena Search | Arena (LMArena) | 1214† | – |
| BALROG | BALROG | 63.4% | 53.2% |
| PostTrainBench | Epoch AI Benchmarking Hub | 35.0%† | – |
| GBAEval | Epoch AI Benchmarking Hub | 79.6%† | – |
| Vending-Bench 2 | Andon Labs | 11181.9† | 7343.2† |
| AutomationBench | Artificial Analysis | 56.6% | 59.6% |
| GDP.pdf (AA) | Artificial Analysis | 21.6% | 24.0% |
| EnterpriseOps-Gym | Artificial Analysis | 47.5% | 38.5% |
| AA-Briefcase v1.1 | Artificial Analysis | 1657 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 28.6% | 13.8% |
| DeepSec-Bench (AA) | Artificial Analysis | 25.9% | – |
| FACTS Search | 85.3%† | 81.7%† | |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 85.6% | 86.0% |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 73.2% | 70.7% |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 98.9% | 99.7% |
| ProofBench | Vals AI | 99.0% | 74.0%† |
| LiveBench Mathematics | LiveBench | 95.7% | 94.9% |
| Arena Maths | Arena (LMArena) | 1517 | 1483† |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 59.9% | 43.2% |
| LiveBench 2.0 Data Analysis | LiveBench | 74.6%† | 79.3%† |
| LiveBench Research | LiveBench | 88.9%† | 60.4%† |
| LiveBench 1.0 Data Analysis | LiveBench | 74.5% | 79.3% |
| AA-Omniscience | Artificial Analysis | 37.1 | 0.1 |
| MMLU-Pro | Vals AI | 91.6%† | 86.7%† |
| LegalBench | Vals AI | 87.0%† | 85.1% |
| CorpFin | Vals AI | 73.2%† | 65.3%† |
| TaxEval | Vals AI | 75.1%† | 76.2%† |
| BioMysteryBench | Vals AI | 79.3%† | – |
| Excel Modeling Benchmark | Vals AI | 73.6%† | 66.2% |
| MedCode | Vals AI | 63.6%† | 43.4%† |
| MedScribe | Vals AI | 91.0%† | 82.9%† |
| Public Benefits Bench v1.1 | Vals AI | 76.9%† | 62.4% |
| DrugDiscoveryBench | Scale AI SEAL | 53.7% | – |
| EBR-bench | Epoch AI Benchmarking Hub | 45.7% | – |
| GDP.pdf | Epoch AI Benchmarking Hub | 24.0% | 24.7%† |
| AA-Omniscience: accuracy | Artificial Analysis | 60.9% | 46.8% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 39.2% | 12.1% |
| Instruction following | |||
| LiveBench Language | LiveBench | 88.7% | 82.9% |
| LiveBench Instruction Following | LiveBench | 63.8% | 64.6% |
| IFBench | Artificial Analysis | – | 71.2% |
| Arena Instruction Following | Arena (LMArena) | 1495 | 1464† |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1491 | 1467† |
| EQ-Bench 4 | EQ-Bench | 1360† | 1224† |
| Arena Creative Writing | Arena (LMArena) | 1474 | 1429† |
| Arena Multi-turn | Arena (LMArena) | 1487 | 1472† |
| Arena Longer Queries | Arena (LMArena) | 1503 | 1468† |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1320† | 1271† |
| LiveBench Design | LiveBench | 83.0%† | 76.0%† |
| MMMU-Pro | Artificial Analysis | 84.7% | 80.7% |
| MMMU-Pro (Vals) | Vals AI | 89.9%† | 86.5% |
| MortgageTax | Vals AI | 72.1%† | 67.3%† |
| SAGE | Vals AI | 49.4%† | 47.0% |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 30.4%† | 30.8%† |
| Long context | |||
| AA-LCR | Artificial Analysis | 79.3% | 83.0% |
| Arena Document | Arena (LMArena) | 1490† | 1477† |
| MLCR | Artificial Analysis | 11.1% | 27.2% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 163.9† | 160.1† |
| LiveBench 2.0 | LiveBench | 82.1%† | 75.0%† |
| LiveBench 1.0 | LiveBench | 80.1% | 77.9% |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 50.8 | 42.1 |
| Vals RSI Index | Vals AI | 33.0% | – |
| Vals Science Index | Vals AI | 47.9%† | – |
| Vals Multimodal Index | Vals AI | 73.9%† | 65.1%† |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | 16.8%† | – |
Data as of 2026-10-25. Best configuration of each model; every score links to its source on the model pages.
What changed
- Claude Opus 5 (max): 1 new result (Vals RSI Index).
- GPT-5.6 Terra (max): 1 new result (Harvey's Legal Agent Benchmark v1).
- Claude Opus 5 (max): 1 score revision.
- GPT-5.6 Terra (max): 1 score revision.
- Claude Opus 5 (max): 1 new result (ForecastBench); 1 score revision.
- Claude Opus 5 (max): BenchLeader Index 68.3 → 68.9.
- GPT-5.6 Terra (max): BenchLeader Index 63.8 → 64.4.
- Claude Opus 5 (max): 4 new results (GDPval-AA v2.1, AA-Omniscience: accuracy, AA-Omniscience: non-hallucination…).
- GPT-5.6 Terra (max): 5 new results (AutomationBench, GDP.pdf (AA), MLCR…).
- Claude Opus 5 (max): 1 score revision.
Related comparisons
- Claude Opus 5 vs Claude Opus 5.5
- Claude Opus 5 vs Gemini 4 Argon
- Claude Opus 5 vs Muse Wonder Full
- GPT-5.6 Terra vs Claude Opus 5.5
- GPT-5.6 Terra vs Gemini 4 Argon
- GPT-5.6 Terra vs Muse Wonder Full
- GPT-5.6 Terra vs Claude Fable 5.1
- Claude Fable 5 vs GPT-5.6 Terra
Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs
Claude Opus 5 vs GPT-5.6 Terra: questions
- Is Claude Opus 5 better than GPT-5.6 Terra?
- Claude Opus 5 (max) leads on quality: 68.8 vs 64.4. The BenchLeader Index combines every independent quality benchmark; Claude Opus 5 (max) is ahead overall as of 2026-10-25, but check the category scores for your use.
- Which is cheaper, Claude Opus 5 or GPT-5.6 Terra, and will that change?
- GPT-5.6 Terra is cheaper today: $4.50 against $10.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Claude Opus 5 better than GPT-5.6 Terra for coding?
- Claude Opus 5 scores higher in coding (70 vs 65 on the category index, where 50 is average). On SciCode, the top shared coding benchmark, it is 56.4% for Claude Opus 5 against 55.0% for GPT-5.6 Terra.
- Is Claude Opus 5 better than GPT-5.6 Terra for agents and tool use?
- Claude Opus 5 scores higher in agentic tasks (64 vs 59 on the category index, where 50 is average). On Terminal-Bench, the top shared agents & tools benchmark, it is 51.8% for Claude Opus 5 against 21.5% for GPT-5.6 Terra.
- Which is better for long documents?
- GPT-5.6 Terra scores higher on long-context benchmarks (67 vs 65). Both take about 1.1M tokens: Claude Opus 5 1,000,000, GPT-5.6 Terra 1,050,000.
- Which has lower latency?
- Claude Opus 5 (max) answers first: 24.1 s against 101.0 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Which is faster, Claude Opus 5 or GPT-5.6 Terra?
- GPT-5.6 Terra streams faster: 126 against 57 output tokens per second.
- Do independent benchmarks agree on Claude Opus 5 vs GPT-5.6 Terra?
- Across the 41 benchmarks both have, Claude Opus 5 leads 31, GPT-5.6 Terra leads 10. Every overall index that rates both puts Claude Opus 5 ahead.
- When was Claude Opus 5 released?
- Claude Opus 5 was released on 24 Jul 2026.