Claude Opus 5 vs GPT-6 Astra
Full profiles: Claude Opus 5 (Anthropic) and GPT-6 Astra (OpenAI): every benchmark, price and rank.
Updated · Data from 11 sources
As of 24 Oct 2026, Claude Opus 5 (max) and GPT-6 Astra (max) are neck and neck on quality (68.8 vs 70.4 on the BenchLeader Index, inside the ±4.7 margin of error), and Claude Opus 5 (max) costs 2.0× less, which makes it the better buy for most workloads. Pick Claude Opus 5 for multimodal (68 vs 66); pick GPT-6 Astra for knowledge (80 vs 72).
Claude Opus 5 benchmarks 
GPT-6 Astra benchmarks 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 | GPT-6 Astra | 74 vs 70 category score; Arena WebDev 1775 vs 1690 | Medium |
| Agents & tool use | Either | 64 vs 64 category score; Terminal-Bench 51.8% vs 58.2% | Either |
| Maths | GPT-6 Astra | 72 vs 71 category score; FrontierMath Tiers 1–3 93.7% vs 85.6% | Medium |
| Reasoning | GPT-6 Astra | 76 vs 74 category score; GPQA Diamond 95.8% vs 93.9% | Medium |
| Knowledge | GPT-6 Astra | 80 vs 72 category score; SimpleQA Verified 75.6% vs 59.9% | High |
| Long documents | Either | 65 vs 65 category score; AA-LCR 79.3% vs 80.7% | Either |
| Images and other media | Claude Opus 5 | 68 vs 66 category score; MMMU-Pro 84.7% vs 86.9% | Medium |
| Chat quality (human preference) | Claude Opus 5 | 69 vs 67 on Arena votes | Medium |
| Lowest cost | Claude Opus 5 | $10.00 vs $20.00 per 1M blended; agentic coding session $0.400 vs $0.800 | – |
| Fastest first answer | Claude Opus 5 | 20.2 s (Claude Opus 5 (max)) vs 247.0 s (GPT-6 Astra (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) and GPT-6 Astra (max) are neck and neck on quality (68.8 vs 70.4, within the ±4.7 margin).
- Claude Opus 5 (max) is stronger in human preference, multimodal.
- GPT-6 Astra (max) is stronger in agents & tools, coding, composite, knowledge, long context, maths, reasoning.
- Claude Opus 5 (max) is 2.0× cheaper ($10.00 vs $20.00 per 1M blended).
At a glance
| Spec | Claude Opus 5 | GPT-6 Astra |
|---|---|---|
| Developer | Anthropic | OpenAI |
| Released | 24 Jul 2026 | 2 Sept 2026 |
| Context window (tokens) | 1,000,000 | 1,050,000 |
| Compared at | Claude Opus 5 (max) | GPT-6 Astra (max) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $5.00 / $25.00 / $0.500 | $10.00 / $50.00 / $1.00 |
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-6 Astra (max) | Leader |
|---|---|---|---|
| Reasoning | 74.1 | 76.2 | GPT-6 Astra |
| Coding | 70.2 | 74.0 | GPT-6 Astra |
| Agents & tools | 64.1 | 64.2 | level |
| Maths | 70.6 | 71.8 | GPT-6 Astra |
| Knowledge | 71.6 | 79.5 | GPT-6 Astra |
| Human preference | 68.5 | 66.7 | Claude Opus 5 |
| Multimodal | 67.9 | 66.0 | Claude Opus 5 |
| Long context | 64.7 | 65.4 | level |
| Composite | 86.3 | 88.4 | GPT-6 Astra |
Where sources disagree
Across the 56 benchmarks both have, Claude Opus 5 leads 16, GPT-6 Astra leads 39 and 1 is a tie. Independent sources split: Claude Opus 5 leads on Arena Text; GPT-6 Astra leads on AA Intelligence Index v4.3.2.
| Overall index | Claude Opus 5 (max) | GPT-6 Astra (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 68.8 | 70.4 | neck and neck |
| AA Intelligence Index v4.3.2 | 50.8 | 52.7 | GPT-6 Astra |
| Arena Text | 1491 | 1477 | 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-6 Astra (max) |
|---|---|---|
| Blended price per 1M tokens | $10.00 | $20.00 |
| Input / output per 1M | $5.00 / $25.00 | $10.00 / $50.00 |
| Cached input per 1M | $0.500 | $1.00 |
| Price per task: chat reply | $0.0095 | $0.019 |
| Price per task: summarize a 30-page report | $0.075 | $0.150 |
| Price per task: code edit | $0.068 | $0.135 |
| Price per task: agentic coding session | $0.400 | $0.800 |
| Price per task: structured extraction | $0.015 | $0.030 |
| Cost to run a full benchmark suite | $5.86 | $3.26 |
| Output speed | 57 tok/s | 58 tok/s |
| Time to first answer | 20.2 s | 247.0 s |
| Time to a complete answer | 29.1 s | 255.6 s |
Per token is not per task: on Artificial Analysis' full suite Claude Opus 5 (max) cost $5.86 and GPT-6 Astra (max) $3.26, because Claude Opus 5 (max) used about 3.6× 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 20 s to its first answer token; Claude Opus 5 (high) scores 67.9 in 9 s, Claude Opus 5 (xhigh) scores 67.4 in 24 s. GPT-6 Astra (max) scores 70.4 and takes 247 s to its first answer token; GPT-6 Astra (high) scores 69.6 in 35 s, GPT-6 Astra (xhigh) scores 68.9 in 130 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
153 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-6 Astra (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 68.8 | 70.4 |
| Agents & tools score | BenchLeader | 64.1 | 64.2 |
| Coding score | BenchLeader | 70.2 | 74.0 |
| Composite score | BenchLeader | 86.3 | 88.4 |
| Human preference score | BenchLeader | 68.5 | 66.7 |
| Knowledge score | BenchLeader | 71.6 | 79.5 |
| Long context score | BenchLeader | 64.7 | 65.4 |
| Maths score | BenchLeader | 70.6 | 71.8 |
| Multimodal score | BenchLeader | 67.9 | 66.0 |
| Reasoning score | BenchLeader | 74.1 | 76.2 |
| Blended price $/M | BenchLeader | $10.00 | $20.00 |
| Output speed | BenchLeader | 57 tok/s | 58 tok/s |
| Time to first answer | BenchLeader | 20.2 s | 247.0 s |
| Context window | BenchLeader | 1,000,000 tokens | 1,050,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 93.9% | 95.8% |
| Humanity's Last Exam | Scale AI / CAIS | – | 54.8%† |
| SimpleBench | SimpleBench | 80.6%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1515 | 1501 |
| LiveBench 2.0 Reasoning | LiveBench | 91.2%† | 92.7%† |
| LiveBench 1.0 Reasoning | LiveBench | 91.2% | 92.7% |
| GPQA Diamond (AA) | Artificial Analysis | 93.2% | 96.1% |
| Humanity's Last Exam (AA) | Artificial Analysis | 54.9% | 54.7% |
| GPQA Diamond (Vals) | Vals AI | 93.4%† | – |
| ARC-AGI-1 | ARC Prize | 97.5% | 97.5% |
| ARC-AGI-2 | ARC Prize | 90.4% | 95.0% |
| ARC-AGI-3 | ARC Prize | 30.2%† | 98.6% |
| CritPt | Artificial Analysis | 29.1% | 31.7% |
| MysteryMechanism | Vals AI | 37.4%† | 53.1% |
| Chess Puzzles | Epoch AI Benchmarking Hub | 43.0% | 74.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 59.0% | 84.0% |
| LMCA | Epoch AI Benchmarking Hub | 63.3% | 64.1%† |
| DTBench | Epoch AI Benchmarking Hub | 97.6% | 97.1%† |
| ForecastBench | ForecastBench | 66.5% | – |
| BTF-3 | Epoch AI Benchmarking Hub | 12.0%† | 13.5%† |
| Coding | |||
| SciCode | SciCode | 56.4% | 56.5% |
| WeirdML | WeirdML | 91.8% | 93.3% |
| FrontierCode | Cognition | 53.4% | 53.3% |
| GSO-Bench | GSO-Bench | – | 79.4%† |
| Arena Coding | Arena (LMArena) | 1528 | 1543 |
Show all 153 benchmarks
| Metric | Publisher | Claude Opus 5 (max) | GPT-6 Astra (max) |
|---|---|---|---|
| Coding | |||
| Arena WebDev | Arena (LMArena) | 1690 | 1775 |
| LiveBench 2.0 Coding | LiveBench | 90.8%† | 93.7%† |
| LiveBench 1.0 Coding | LiveBench | 81.5% | 80.4% |
| LiveCodeBench | Vals AI | 89.0%† | – |
| IOI | Vals AI | 84.3%† | 100.0% |
| SWE-bench (Vals) | Vals AI | 97.0%† | – |
| Code Migration | Vals AI | 57.5%† | 67.7% |
| Code Migration v1.1 | Vals AI | – | 80.6% |
| ProgramBench | Vals AI | 3.0%† | 5.5% |
| SWE-bench Multimodal | Vals AI | – | 43.3% |
| Vibe Code Bench 1-100 | Vals AI | 28.5%† | 27.6% |
| Vibe Code Bench v1.1 | Vals AI | 88.4%† | 89.6% |
| SWE Atlas: Codebase QnA | Scale AI SEAL | 63.2%† | 59.1%† |
| SWE Atlas: Refactoring | Scale AI SEAL | – | 59.0%† |
| SWE Atlas: Test Writing | Scale AI SEAL | 62.2%† | 50.7%† |
| MirrorCode | Epoch AI Benchmarking Hub | – | 46.7%† |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 73.7% | 73.2% |
| CursorBench | Cursor | 46.6% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 2164.6† | 2951.3 |
| FrontierSWE | Epoch AI Benchmarking Hub | 52.0% | 65.5% |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 49.0% | 59.1% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 89.1% | 88.4% |
| Startrise Frontend | Startrise | 82.3%† | 79.9%† |
| FrontierCode 1.1 (AA) | Artificial Analysis | – | 47.0% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 51.8% | 58.2% |
| OSWorld-Verified 2.0 | OSWorld | 31.4% | – |
| Remote Labor Index | Scale AI / CAIS | – | 20.8%† |
| APEX-Agents | Mercor | 65.8% | 64.7%† |
| Arena Agent | Arena (LMArena) | 8.1 | 14.7 |
| LiveBench Long Agentic | LiveBench | 62.1%† | 82.4%† |
| LiveBench Browser | LiveBench | 84.0%† | 84.0%† |
| LiveBench Agentic Coding | LiveBench | 65.2% | 57.3% |
| Terminal-Bench 2.1 (Vals) | Vals AI | 84.6%† | 87.3% |
| MCP Atlas | Scale AI SEAL | 85.8%† | – |
| HiL-Bench | Scale AI SEAL | 57.0%† | 65.3%† |
| τ³-bench Banking | τ²-bench | 48.7% | – |
| GDPval-AA v2.1 | Artificial Analysis | – | 53.7% |
| τ³-Banking (AA) | Artificial Analysis | 42.1% | 41.4% |
| ITBench SRE (AA) | Artificial Analysis | – | 48.6% |
| Analyst Agent (AA) | Artificial Analysis | 53.8% | 51.3% |
| CUA-bench | Vals AI | 9.0% | 19.2% |
| CyberBench | Vals AI | 65.4%† | 41.1% |
| Finance Agent v2 | Vals AI | 58.6%† | 53.5% |
| Finance Agent v2.1 | Vals AI | – | 60.4% |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | – | 7.1% |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 6.7%† | 5.4% |
| Legal Research Bench | Vals AI | 55.3%† | 39.4% |
| Robo-bench | Vals AI | – | 23.3% |
| SkillsBench | Vals AI | 60.4%† | – |
| SREBench | Vals AI | 12.2%† | 56.9% |
| Tax Agent Bench | Vals AI | 45.5%† | 20.7% |
| Terminal-Bench 4.0 (Vals) | Vals AI | 53.5%† | 59.6% |
| Terminal-Bench Science | Vals AI | 27.1%† | 62.9% |
| Time Horizon Index: KSP | Vals AI | 18.8%† | 90.5% |
| Arena Search | Arena (LMArena) | 1214† | 1246 |
| BALROG | BALROG | 63.4% | 68.3% |
| PostTrainBench | Epoch AI Benchmarking Hub | 35.0%† | – |
| GBAEval | Epoch AI Benchmarking Hub | 79.6%† | – |
| Vending-Bench 2 | Andon Labs | 11181.9† | 15514.7† |
| AutomationBench | Artificial Analysis | 56.6% | 68.5% |
| GDP.pdf (AA) | Artificial Analysis | 21.6% | 31.0% |
| Harvey LAB-AA v1.1 | Harvey | – | 8.6% |
| EnterpriseOps-Gym | Artificial Analysis | 47.5% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1657 | 1568 |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 28.6% | 63.3% |
| CWE-Bench (AA) | Artificial Analysis | – | 63.3% |
| DeepSec-Bench (AA) | Artificial Analysis | 25.9% | 36.9% |
| CyberGym end-to-end (AA) | Artificial Analysis | – | 0.0% |
| FACTS Search | 85.3%† | 89.9%† | |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 85.6% | 93.7% |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 73.2% | 97.6% |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 98.9% | 100.0% |
| ProofBench | Vals AI | 99.0% | 99.0%† |
| LiveBench Mathematics | LiveBench | 95.7% | 96.8% |
| Arena Maths | Arena (LMArena) | 1517 | 1487 |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 59.9% | 75.6% |
| LiveBench 2.0 Data Analysis | LiveBench | 74.6%† | 83.0%† |
| LiveBench Research | LiveBench | 88.9%† | 85.2%† |
| LiveBench 1.0 Data Analysis | LiveBench | 74.5% | 83.0% |
| AA-Omniscience | Artificial Analysis | 37.1 | 43.4 |
| MMLU-Pro | Vals AI | 91.6%† | – |
| LegalBench | Vals AI | 87.0%† | – |
| CorpFin | Vals AI | 73.2%† | – |
| TaxEval | Vals AI | 75.1%† | – |
| PRBench Finance | Scale AI SEAL | – | 47.5%† |
| PRBench Legal | Scale AI SEAL | – | 48.4%† |
| BioMysteryBench | Vals AI | 79.3%† | 79.3% |
| Excel Modeling Benchmark | Vals AI | 73.6%† | 71.7% |
| Excel Modeling Benchmark v1.1 | Vals AI | – | 72.7% |
| MedCode | Vals AI | 63.6%† | 48.5% |
| MedScribe | Vals AI | 91.0%† | 87.9% |
| Public Benefits Bench v1.1 | Vals AI | 76.9%† | – |
| DrugDiscoveryBench | Scale AI SEAL | 53.7% | 68.7% |
| EBR-bench | Epoch AI Benchmarking Hub | 45.7% | 76.2% |
| GDP.pdf | Epoch AI Benchmarking Hub | 24.0% | 34.2% |
| AA-Omniscience: accuracy | Artificial Analysis | 60.9% | 62.6% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 39.2% | 48.7% |
| Instruction following | |||
| LiveBench Language | LiveBench | 88.7% | 89.4% |
| LiveBench Instruction Following | LiveBench | 63.8% | 75.6% |
| Arena Instruction Following | Arena (LMArena) | 1495 | 1473 |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1491 | 1477 |
| EQ-Bench 4 | EQ-Bench | 1385† | – |
| Arena Creative Writing | Arena (LMArena) | 1474 | 1450 |
| Arena Multi-turn | Arena (LMArena) | 1487 | 1485 |
| Arena Longer Queries | Arena (LMArena) | 1503 | 1489 |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1320† | 1280 |
| LiveBench Design | LiveBench | 83.0%† | 82.0%† |
| MMMU-Pro | Artificial Analysis | 84.7% | 86.9% |
| MMMU-Pro (Vals) | Vals AI | 89.9%† | – |
| MortgageTax | Vals AI | 72.1%† | – |
| SAGE | Vals AI | 49.4%† | 46.4% |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 30.4%† | 49.7%† |
| AA-3D | Artificial Analysis | – | 1166 |
| Long context | |||
| AA-LCR | Artificial Analysis | 79.3% | 80.7% |
| Arena Document | Arena (LMArena) | 1490† | 1473 |
| MLCR | Artificial Analysis | 11.1% | 41.7% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 163.9† | 168.8† |
| LiveBench 2.0 | LiveBench | 82.1%† | 86.1%† |
| LiveBench 1.0 | LiveBench | 80.1% | 82.2% |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 50.8 | 52.7 |
| Vals Index | Vals AI | – | 65.0 |
| Vals Finance Index | Vals AI | – | 55.6% |
| Vals RSI Index | Vals AI | 33.0% | 27.1% |
| Vals Science Index | Vals AI | 47.9%† | 65.1% |
| Vals Multimodal Index | Vals AI | 73.9%† | – |
| Vals Router Index | Vals AI | – | 65.0%† |
| AA Cyber Index | Artificial Analysis | – | 33.4 |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | 16.8%† | – |
Data as of 2026-10-24. 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-6 Astra (max): 9 new results (Code Migration v1.1, Excel Modeling Benchmark v1.1, Finance Agent v2.1…).
- Claude Opus 5 (max): 1 score revision.
- GPT-6 Astra (max): 1 score revision.
- Claude Opus 5 (max): 1 new result (ForecastBench); 1 score revision.
- GPT-6 Astra (max): 4 score revisions.
- Claude Opus 5 (max): BenchLeader Index 68.3 → 68.9.
- GPT-6 Astra (max): BenchLeader Index 69.8 → 70.6.
- Claude Opus 5 (max): 4 new results (GDPval-AA v2.1, AA-Omniscience: accuracy, AA-Omniscience: non-hallucination…).
- 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-6 Astra vs Claude Opus 5.5
- GPT-6 Astra vs Gemini 4 Argon
- GPT-6 Astra vs Muse Wonder Full
- GPT-6 Astra vs Claude Fable 5.1
- Claude Sonnet 5 vs GPT-6 Astra
Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs
Claude Opus 5 vs GPT-6 Astra: questions
- Is Claude Opus 5 better than GPT-6 Astra?
- Claude Opus 5 (max) and GPT-6 Astra (max) are neck and neck on quality (68.8 vs 70.4, within the ±4.7 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-24, so check the category scores for your use.
- Which is cheaper, Claude Opus 5 or GPT-6 Astra, and will that change?
- Claude Opus 5 is cheaper today: $10.00 against $20.00 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-6 Astra for coding?
- GPT-6 Astra scores higher in coding (74 vs 70 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1690 for Claude Opus 5 against 1775 for GPT-6 Astra.
- Is Claude Opus 5 better than GPT-6 Astra for agents and tool use?
- They are level in agentic tasks (64 vs 64 on the category index). On Terminal-Bench, the top shared agents & tools benchmark, it is 51.8% for Claude Opus 5 against 58.2% for GPT-6 Astra.
- Which is better for long documents?
- They are level on long-context benchmarks (65 vs 65). Both take about 1.1M tokens: Claude Opus 5 1,000,000, GPT-6 Astra 1,050,000.
- Which has lower latency?
- Claude Opus 5 (max) answers first: 20.2 s against 247.0 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-6 Astra?
- GPT-6 Astra streams faster: 58 against 57 output tokens per second.
- Do independent benchmarks agree on Claude Opus 5 vs GPT-6 Astra?
- Across the 56 benchmarks both have, Claude Opus 5 leads 16, GPT-6 Astra leads 39 and 1 is a tie. Independent sources split: Claude Opus 5 leads on Arena Text; GPT-6 Astra leads on AA Intelligence Index v4.3.2.
- When was GPT-6 Astra released?
- GPT-6 Astra was released on 2 Sept 2026.