Claude Opus 4.6 vs GPT-5
Full profiles: Claude Opus 4.6 (Anthropic) and GPT-5 (OpenAI): every benchmark, price and rank.
Updated · Data from 11 sources
As of 16 Nov 2026, Claude Opus 4.6 and GPT-5 are neck and neck on quality (62.5 vs 59.0 on the BenchLeader Index, inside the ±6.0 margin), and GPT-5 is 2.9× cheaper at $3.44 vs $10.00 per 1M tokens blended. Choose Claude Opus 4.6 for reasoning (65 vs 48), multimodal (67 vs 57), and coding (62 vs 57) and fast first answers (1.1 s vs 22.9 s); choose GPT-5 for maths (65 vs 63), where it also streams ~2× faster.
Key facts (as of 16 Nov 2026)
- Claude Opus 4.6 scores 62.5 on the BenchLeader Index and GPT-5 scores 59.0 (BenchLeader, 16 Nov 2026).
- Claude Opus 4.6 costs $10.00 per 1M tokens blended (list price, 16 Nov 2026).
- GPT-5 costs $3.44 per 1M tokens blended (list price, 16 Nov 2026).
- GPT-5 streams ~2× faster than Claude Opus 4.6 (Artificial Analysis, 7-day median, 16 Nov 2026).
- Of the 4 benchmarks both models have, Claude Opus 4.6 leads 4 (BenchLeader, 16 Nov 2026).
- On Arena WebDev, Claude Opus 4.6 is rated 1547 and GPT-5 1417 (LMArena, as published by 16 Nov 2026).
Which 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 4.6 | 62 vs 57 category score; Arena WebDev 1547 vs 1417 | Medium |
| Maths | GPT-5 | 65 vs 63 category score | Medium |
| Reasoning | Claude Opus 4.6 | 65 vs 48 category score; ARC-AGI-2 69.2% vs 7.5% | High |
| Images and other media | Claude Opus 4.6 | 67 vs 57 category score | High |
| Chat quality (human preference) | – | Only Claude Opus 4.6 has an Arena rating (70 category score) | Not comparable |
| Lowest cost | GPT-5 | $3.44 vs $10.00 per 1M blended; agentic coding session $0.115 vs $0.400 | – |
| Fastest first answer | Claude Opus 4.6 | 1.1 s (Claude Opus 4.6 (high)) vs 22.9 s (GPT-5 (medium)) 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 4.6 (high) and GPT-5 (medium) are neck and neck on quality (62.5 vs 59.0, within the ±6.0 margin).
- Claude Opus 4.6 (high) is stronger in coding, multimodal, reasoning.
- GPT-5 (medium) is stronger in maths.
- GPT-5 (medium) is 2.9× cheaper ($3.44 vs $10.00 per 1M blended).
- GPT-5 (medium) streams 2.3× faster (82 vs 35 tokens per second).
At a glance
| Spec | Claude Opus 4.6 | GPT-5 |
|---|---|---|
| Developer | Anthropic | OpenAI |
| Released | 5 Feb 2026 | 7 Aug 2025 |
| Context window (tokens) | 1,000,000 | 400,000 |
| Compared at | Claude Opus 4.6 (high) | GPT-5 (medium) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $5.00 / $25.00 / $0.500 | $1.25 / $10.00 / $0.125 |
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 4.6 (high) | GPT-5 (medium) | Leader |
|---|---|---|---|
| Reasoning | 65.0 | 47.9 | Claude Opus 4.6 |
| Coding | 62.2 | 57.4 | Claude Opus 4.6 |
| Agents & tools | – | 58.1 | not comparable |
| Maths | 63.4 | 65.4 | GPT-5 |
| Instruction following | – | 69.4 | not comparable |
| Human preference | 69.8 | – | not comparable |
| Multimodal | 66.7 | 56.6 | Claude Opus 4.6 |
| Long context | – | 70.4 | not comparable |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Not drawn, measured for fewer than two of these models: Agents & tools, Knowledge, Instruction following, Long context, Human preference.
Where sources disagree
Across the 4 benchmarks both have, Claude Opus 4.6 leads 4, GPT-5 leads 0. Every overall index that rates both puts Claude Opus 4.6 ahead.
| Overall index | Claude Opus 4.6 (high) | GPT-5 (medium) | Leader |
|---|---|---|---|
| BenchLeader Index | 62.5 | 59.0 | neck and neck |
| Arena Text | 1503 | 1435† | Claude Opus 4.6 |
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. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | Claude Opus 4.6 (high) | GPT-5 (medium) |
|---|---|---|
| Blended price per 1M tokens | $10.00 | $3.44 |
| Input / output per 1M | $5.00 / $25.00 | $1.25 / $10.00 |
| Cached input per 1M | $0.500 | $0.125 |
| Price per task: chat reply | $0.0095 | $0.0035 |
| Price per task: summarize a 30-page report | $0.075 | $0.021 |
| Price per task: code edit | $0.068 | $0.022 |
| Price per task: agentic coding session | $0.400 | $0.115 |
| Price per task: structured extraction | $0.015 | $0.0045 |
| Agentic coding session, input mostly cached | $0.198 | $0.064 |
| Output speed | 35 tok/s | 82 tok/s |
| Time to first answer | 1.1 s | 22.9 s |
| Time to a complete answer, last measured 13 Nov 2026 | – | 66.2 s |
Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 4.6 (high) scores 62.5 and takes 1 s to its first answer token; Claude Opus 4.6 scores 62.1 in 1 s, Claude Opus 4.6 (no reasoning) scores 58.4 in 2 s. GPT-5 (medium) scores 59.0 and takes 23 s to its first answer token; GPT-5 (high) scores 58.2 in 49 s, GPT-5 scores 56.6 in 5 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
145 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 4.6 (high) | GPT-5 (medium) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 62.5 | 59.0 |
| Coding score | BenchLeader | 62.2 | 57.4 |
| Human preference score | BenchLeader | 69.8 | – |
| Maths score | BenchLeader | 63.4 | 65.4 |
| Multimodal score | BenchLeader | 66.7 | 56.6 |
| Reasoning score | BenchLeader | 65.0 | 47.9 |
| Agents & tools score | BenchLeader | – | 58.1 |
| Instruction following score | BenchLeader | – | 69.4 |
| Long context score | BenchLeader | – | 70.4 |
| Blended price $/M | BenchLeader | $10.00 | $3.44 |
| Output speed | BenchLeader | 35 tok/s | 82 tok/s |
| Time to first answer | BenchLeader | 1.1 s | 22.9 s |
| Context window | BenchLeader | 1,000,000 tokens | 400,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 90.5%† | 85.3% |
| Humanity's Last Exam | Scale AI / CAIS | 19.0%† | 25.3%† |
| SimpleBench | SimpleBench | 67.6%† | 56.7%† |
| Arena Hard Prompts | Arena (LMArena) | 1530 | 1449† |
| LiveBench 1.0 Reasoning | LiveBench | 88.7% | – |
| GPQA Diamond (AA) | Artificial Analysis | 84.0%† | 84.2% |
| Humanity's Last Exam (AA) | Artificial Analysis | 19.1%† | 25.4% |
| GPQA Diamond (Vals) | Vals AI | 89.7%† | 85.6%† |
| Kagi LLM Benchmark | Kagi LLM Benchmark | 72.4%† | 72.7%† |
| GPQA Diamond (HELM) | HELM Capabilities | – | 79.1%† |
| ARC-AGI-1 | ARC Prize | 94.0% | 56.2% |
| ARC-AGI-2 | ARC Prize | 69.2% | 7.5% |
| ARC-AGI-3 | ARC Prize | 0.5%† | – |
| CritPt | Artificial Analysis | 2.8%† | 0.0% |
| CritPt v2 hard subset (AA) | Artificial Analysis | – | 11.5%† |
| Arena Expert | Arena (LMArena) | 1544 | 1445† |
| Chess Puzzles | Epoch AI Benchmarking Hub | 21.0%† | 32.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 15.0%† | 15.0% |
| CL-bench | Epoch AI Benchmarking Hub | 20.7%† | – |
| CL-bench Life | Epoch AI Benchmarking Hub | 17.0% | – |
| LMCA | Epoch AI Benchmarking Hub | 55.8%† | 40.0%† |
| DTBench | Epoch AI Benchmarking Hub | 91.2%† | 90.7%† |
| ForecastBench | ForecastBench | 63.1%† | 65.0%† |
| Coding | |||
| SWE-bench Verified (Epoch) | Epoch AI Benchmarking Hub | 78.7%† | 71.5% |
| SciCode | SciCode | – | 42.9%† |
Show all 145 benchmarks
| Metric | Publisher | Claude Opus 4.6 (high) | GPT-5 (medium) |
|---|---|---|---|
| Coding | |||
| WeirdML | WeirdML | 78.0% | 39.8%† |
| FrontierCode | Cognition | 26.6%† | – |
| GSO-Bench | GSO-Bench | 41.2% | 6.9%† |
| Arena Coding | Arena (LMArena) | 1547 | 1462† |
| Arena WebDev | Arena (LMArena) | 1547 | 1417 |
| LiveBench 1.0 Coding | LiveBench | 78.2% | – |
| LiveCodeBench | Vals AI | 84.7%† | 85.9%† |
| SWE-bench (Vals) | Vals AI | 78.2%† | 69.0%† |
| SWE-Bench Pro | Scale AI SEAL | 51.9%† | 41.8%† |
| Aider Polyglot | Aider polyglot leaderboard | – | 86.7% |
| SWE-bench Verified (bash only) | SWE-bench | 75.6%† | 65.0% |
| SWE-bench Multilingual (bash only) | SWE-bench | 72.0%† | – |
| SWE-bench Verified (any scaffold) | SWE-bench | 75.6%† | 74.4%† |
| Vibe Code Bench v1.1 | Vals AI | 57.6%† | 20.1%† |
| SWE-Bench Pro (private) | Scale AI SEAL | 47.1%† | 14.9%† |
| SWE Atlas v1.0: Codebase QnA | Scale AI SEAL | 33.3%† | – |
| SWE Atlas v1.0: Refactoring | Scale AI SEAL | 35.6%† | – |
| SWE Atlas v1.0: Test Writing | Scale AI SEAL | 36.7%† | – |
| MirrorCode | Epoch AI Benchmarking Hub | 33.3%† | 4.4%† |
| ALE-Bench | Epoch AI Benchmarking Hub | 996.5† | 1162.5† |
| AlgoTune | AlgoTune | 1.5 | 1.7† |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | – | 35.2%† |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | – | 38.0 |
| Design Arena Website | Design Arena | 1283† | 1174† |
| LiveCodeBench (AA) | Artificial Analysis | – | 70.3% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 79.8%† | 49.6% |
| GDPval | OpenAI | – | 34.8% |
| Cybench | Cybench | 93.0%† | – |
| Remote Labor Index | Scale AI / CAIS | 4.2%† | 1.7%† |
| APEX-Agents | Mercor | 46.3%† | – |
| LiveBench Agentic Coding | LiveBench | 49.0% | – |
| Terminal-Bench Hard | Artificial Analysis | 48.5%† | 37.9% |
| τ²-Bench Telecom (AA) | Artificial Analysis | 84.8%† | 86.5% |
| MCP Atlas | Scale AI SEAL | 76.8%† | – |
| HiL-Bench | Scale AI SEAL | 38.3%† | – |
| τ³-bench Banking | τ²-bench | 27.3%† | – |
| GDPval-AA v2.1 | Artificial Analysis | – | 925† |
| τ³-Banking (AA) | Artificial Analysis | – | 22.1%† |
| APEX-Agents (AA) | Artificial Analysis | 33.0%† | – |
| Poker Agent | Vals AI | – | 1103.2%† |
| Terminal-Bench 2.0 (Vals) | Vals AI | 58.4%† | 37.1%† |
| Arena Search | Arena (LMArena) | 1222† | 1189† |
| BALROG | BALROG | – | 32.8%† |
| DeepResearch Bench | DeepResearch Bench | 55.3% | 48.6% |
| METR Time Horizons | METR | 78.9%† | 69.6% |
| GBAEval | Epoch AI Benchmarking Hub | 44.1%† | – |
| Vending-Bench 2 | Andon Labs | 8017.6† | – |
| Design Arena Agentic | Design Arena | 1196† | – |
| BrowseComp-Plus | BrowseComp-Plus | – | 57.6%† |
| BrowseComp | OpenAI | – | 20.1%† |
| DeepSearchQA | – | 59.4%† | |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 66.0%† | 55.4%† |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 26.8%† | 21.9%† |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 94.4%† | 87.2% |
| MATH Level 5 | Epoch AI Benchmarking Hub | – | 97.9% |
| ProofBench | Vals AI | 50.0%† | 18.0%† |
| LiveBench Mathematics | LiveBench | 89.3% | – |
| AIME (Vals) | Vals AI | 95.6%† | 93.4%† |
| MGSM | Vals AI | – | 92.8%† |
| AIME 2026 | MathArena | 96.7% | – |
| HMMT February 2026 | MathArena | 96.2% | – |
| USAMO 2026 | MathArena | 47.0% | – |
| IMO 2025 | MathArena | – | 38.1%† |
| MathArena Apex | MathArena | 34.5% | 1.0%† |
| Omni-MATH (HELM) | HELM Capabilities | – | 64.7%† |
| MATH 500 | Vals AI | – | 96.0%† |
| Arena Maths | Arena (LMArena) | 1516 | 1411† |
| AIME 2025 (AA) | Artificial Analysis | – | 91.7% |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 47.0%† | 50.1%† |
| LiveBench 1.0 Data Analysis | LiveBench | 69.9% | – |
| MMLU-Pro | Vals AI | 89.1%† | 86.5%† |
| LegalBench | Vals AI | 85.3%† | 86.0%† |
| CorpFin | Vals AI | 67.0%† | 61.1%† |
| TaxEval | Vals AI | 76.0%† | 73.4%† |
| MedQA | Vals AI | 95.4%† | 96.3%† |
| PRBench Finance | Scale AI SEAL | 53.3%† | 51.3%† |
| PRBench Legal | Scale AI SEAL | 52.3%† | 49.0%† |
| MultiNRC | Scale AI SEAL | 48.3%† | 52.1%† |
| MMLU-Pro (HELM) | HELM Capabilities | – | 86.3%† |
| CaseLaw v2 | Vals AI | 62.1%† | 66.5%† |
| MedCode | Vals AI | 49.1%† | 49.6%† |
| MedScribe | Vals AI | 86.7%† | 83.7%† |
| DrugDiscoveryBench | Scale AI SEAL | 27.7%† | – |
| EBR-bench | Epoch AI Benchmarking Hub | 12.7%† | 12.7%† |
| Global-MMLU-Lite (AA) | Artificial Analysis | 90.6%† | 90.4% |
| Instruction following | |||
| LiveBench Language | LiveBench | 83.3% | – |
| LiveBench Instruction Following | LiveBench | 63.3% | – |
| IFBench | Artificial Analysis | 44.6%† | 70.6% |
| MultiChallenge | Scale AI SEAL | 56.0%† | 63.2%† |
| TutorBench | Scale AI SEAL | 53.5%† | 55.3%† |
| IFEval (HELM) | HELM Capabilities | – | 87.5%† |
| WildBench (HELM) | HELM Capabilities | – | 85.7%† |
| Arena Instruction Following | Arena (LMArena) | 1511 | 1415† |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1503 | 1427† |
| EQ-Bench 4 | EQ-Bench | 1210† | – |
| EQ-Bench Creative Writing v3 | EQ-Bench | 1810† | 1626† |
| Arena Creative Writing | Arena (LMArena) | 1499 | 1389† |
| Arena Multi-turn | Arena (LMArena) | 1517 | 1449† |
| Arena Longer Queries | Arena (LMArena) | 1522 | 1435† |
| Lech Mazur Creative Writing | Lech Mazur | – | 86.0% |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1316 | 1232† |
| MMMU-Pro | Artificial Analysis | 72.5%† | 74.3% |
| VISTA | Scale AI SEAL | 45.5%† | 49.7%† |
| MMMU-Pro (Vals) | Vals AI | 83.9%† | 81.5%† |
| MortgageTax | Vals AI | 68.5%† | 65.5%† |
| VTB | Scale AI SEAL | 27.5%† | 17.0%† |
| GeoBench | GeoBench | – | 81.0% |
| VPCT | Epoch AI Benchmarking Hub | – | 63.2% |
| Furniture Assembly | Epoch AI Benchmarking Hub | 28.3%† | – |
| Long context | |||
| Fiction.LiveBench 120k | Fiction.live | – | 96.9% |
| AA-LCR v1.1 | Artificial Analysis | 67.0%† | 76.0% |
| Arena Document | Arena (LMArena) | 1497 | – |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 156.3† | 150† |
| LiveBench 1.0 | LiveBench | 74.5% | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 26.4† | 22.9 |
| AA Progress Index v1 | Artificial Analysis | 274.5† | 264.4 |
| HELM Capabilities mean | HELM Capabilities | – | 80.7%† |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | 20.5%† | 17.0%† |
| MASK | Scale AI SEAL | 96.3%† | 79.3%† |
| Metadata | |||
| AA Openness Index | Artificial Analysis | – | 5.6 |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- GPT-5 (medium): BenchLeader Index 58.2 → 59.0.
- GPT-5 (medium): 1 new result (Aider Polyglot).
- GPT-5 (medium): 2 new results (Global-MMLU-Lite (AA), AIME 2025 (AA)); 1 score revision.
- Claude Opus 4.6 (high): BenchLeader Index 63.0 → 62.5.
- GPT-5 (medium): 1 score revision.
- GPT-5 (medium): 1 score revision.
- GPT-5 (medium): 1 score revision.
- GPT-5 (medium): 1 new result (AA Progress Index v1).
- Claude Opus 4.6 (high): 1 new result (USAMO 2026).
Related comparisons
- Claude Opus 4.6 vs Claude Fable 5.2
- Claude Opus 4.6 vs Gemini 4 Carbon
- Claude Opus 4.6 vs Claude Opus 5.5
- Claude Opus 4.6 vs GPT-6 Astra
- GPT-5 vs Claude Opus 5.5
- Claude Opus 4.6 vs Claude Fable 5.1
- Claude Opus 4.6 vs GPT-6.1 Sol
- Claude Opus 4.6 vs Claude Fable 5
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Claude Opus 4.6 vs GPT-5: questions
- Is Claude Opus 4.6 better than GPT-5?
- Claude Opus 4.6 (high) and GPT-5 (medium) are neck and neck on quality (62.5 vs 59.0, within the ±6.0 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
- Claude Opus 4.6 vs GPT-5: which is cheaper, and will that change?
- GPT-5 is cheaper today: $3.44 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 4.6 or GPT-5 better for coding?
- Claude Opus 4.6 scores higher in coding (62 vs 57 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1547 for Claude Opus 4.6 against 1417 for GPT-5.
- Which is faster, Claude Opus 4.6 or GPT-5?
- GPT-5 streams ~2× faster (82 vs 35 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- Claude Opus 4.6 accepts more context: 1,000,000 against 400,000 tokens for GPT-5.
More questions (3)
- Which has lower latency?
- Claude Opus 4.6 answers first: 1.1 s against 22.9 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.
- Do independent benchmarks agree on Claude Opus 4.6 vs GPT-5?
- Across the 4 benchmarks both have, Claude Opus 4.6 leads 4, GPT-5 leads 0. Every overall index that rates both puts Claude Opus 4.6 ahead.
- When was Claude Opus 4.6 released?
- Claude Opus 4.6 was released on 5 Feb 2026.