AnthropicReasoning modelClaude Opus 4.8
Best configuration ranks #31 of 610 on the BenchLeader Index at 64.7 ±3.3. Last measured 8 Sept 2026. Released 28 May 2026.
- Blended price
- $10.00/M
- $5.00 in · $25.00 out
- Output speed
- 58 tok/s
- First answer
- 29 s
- first token 1.68 s
- Context
- 1M
- Overall index65
- Reasoning64
- Coding66
- Agents & tools65
- Knowledge66
- Instruction following62
- Human preference66
- Multimodal64
- Long context65
- Composite82
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | – | – | 58 tok/s | 29 s | $0.0095 | Coding 64 · Maths 61 · Reasoning 62 |
| low | – | – | 58 tok/s | 29 s | $0.0095 | Maths 67 · Reasoning 63 |
| medium | – | – | 58 tok/s | 29 s | $0.0095 | Coding 68 · Reasoning 66 |
| high | 61.3 | #70 | 58 tok/s | 29 s | $0.0095 | Agents & tools 65 · Coding 65 · Human preference 68 · Multimodal 65 · Reasoning 62 |
| xhigh | – | – | 58 tok/s | 29 s | $0.0095 | Coding 73 · Reasoning 30 |
| max | 59.4 | #104 | 58 tok/s | 29 s | $0.0095 | Agents & tools 54 · Coding 62 · Composite 55 · Knowledge 61 · Maths 70 · Reasoning 65 |
| thinking | – | – | 58 tok/s | 29 s | $0.0095 | Reasoning 80 |
| defaultbest | 64.7 | #31 | 58 tok/s | 29 s | $0.0095 | Agents & tools 65 · Coding 66 · Composite 82 · Human preference 66 · Instruction following 62 · Knowledge 66 · Long context 65 · Multimodal 64 · Reasoning 64 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| GPQA Diamond | 85.3%#76 | 88.4%#49 | – | – | – | 91.0%#27 | – | – | Epoch AI Benchmarking Hub | |
| SimpleBench | – | – | – | – | – | – | – | 64.8%#17 | SimpleBench | |
| LMArena Hard Prompts | – | – | – | 1515#10 | – | – | – | 1503#21 | LMArena | |
| LiveBench Reasoningnot in index | – | – | – | – | – | 89.2%#13 | – | – | LiveBench | |
| GPQA Diamond (AA)not in index | – | – | – | – | – | – | – | 92.0%#41 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | – | – | – | – | – | – | 48.7%#18 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | – | – | – | – | – | – | 92.4%#16 | Vals AI | |
| EnigmaEval | – | – | – | – | 23.5%#5 | – | – | – | Scale AI SEAL | |
| Kagi LLM Benchmark | – | – | – | – | – | – | 88.8%#2 | – | Kagi LLM Benchmark | |
| ARC-AGI-1 | – | 88.0%#57 | 91.5%#44 | 92.0%#40 | – | 92.5%#35 | – | – | ARC Prize | |
| ARC-AGI-2 | – | 62.2%#51 | 71.7%#36 | 72.1%#34 | – | – | – | – | ARC Prize | |
| ARC-AGI-3 | – | – | – | 1.5%#18 | – | – | – | – | ARC Prize |
Coding
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| SciCode | – | – | – | – | – | 53.5%#41 | – | – | SciCode | |
| WeirdML | 70.5%#29 | – | 76.0%#24 | – | 82.9%#14 | – | – | – | WeirdML | |
| FrontierCode | – | – | – | – | – | – | – | 46.5%#7 | Cognition | |
| GSO-Bench | – | – | – | – | – | – | – | 47.1%#1 | GSO-Bench | |
| LMArena Coding | – | – | – | 1534#8 | – | – | – | 1527#17 | LMArena | |
| LMArena WebDev | – | – | – | 1561#21 | – | – | – | 1540#27 | LMArena | |
| LiveBench Codingnot in index | – | – | – | – | – | 81.8%#8 | – | – | LiveBench | |
| SciCode (AA)not in index | – | – | – | – | – | – | – | 54.4%#38 | Artificial Analysis | |
| LiveCodeBench | – | – | – | – | – | – | – | 87.8%#11 | Vals AI | |
| SWE-bench (Vals)not in index | – | – | – | – | – | – | – | 88.6%#12 | Vals AI |
Agents & tools
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| Terminal-Bench | – | – | – | – | – | 23.6%#51 | – | – | Terminal-Bench | |
| OSWorld-Verified 2.0 | – | – | – | – | – | 20.6%#7 | – | – | OSWorld | |
| Remote Labor Index | – | – | – | – | – | – | – | 8.3%#2 | Scale AI / CAIS | |
| APEX-Agents | – | – | – | – | – | 42.5%#6 | – | – | Mercor | |
| LMArena Agent | – | – | – | 7.7#6 | – | – | – | – | LMArena | |
| LiveBench Agentic Codingnot in index | – | – | – | – | – | 50.5%#30 | – | – | LiveBench | |
| Terminal-Bench Hard | – | – | – | – | – | – | – | 58.3%#10 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | – | – | – | – | – | – | – | 94.4%#31 | Artificial Analysis | |
| Terminal-Bench 2.1 (Vals) | – | – | – | – | – | – | – | 71.9%#18 | Vals AI | |
| MCP Atlas | – | – | – | – | – | 82.2%#7 | – | – | Scale AI SEAL | |
| HiL-Bench | – | – | – | – | – | – | – | 35.3%#9 | Scale AI SEAL |
Maths
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| FrontierMath Tiers 1–3 | – | – | – | – | – | 80.0%#11 | – | – | Epoch AI Benchmarking Hub | |
| FrontierMath Tier 4 | – | – | – | – | – | 56.1%#18 | – | – | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 84.4%#83 | 97.8%#17 | – | – | – | 98.3%#14 | – | – | Epoch AI Benchmarking Hub | |
| ProofBench | – | – | – | – | – | 69.0%#9 | – | – | Vals AI | |
| LiveBench Mathematicsnot in index | – | – | – | – | – | 94.3%#10 | – | – | LiveBench | |
| AIME 2026 | – | – | – | – | – | 100.0%#1 | – | – | MathArena | |
| HMMT February 2026 | – | – | – | – | – | 95.5%#6 | – | – | MathArena | |
| MathArena Apex | – | – | – | – | – | 81.3%#1 | – | – | MathArena |
Knowledge
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| SimpleQA Verified | – | – | – | – | – | 53.0%#16 | – | – | Epoch AI Benchmarking Hub | |
| LiveBench Data Analysisnot in index | – | – | – | – | – | 66.0%#44 | – | – | LiveBench | |
| AA-Omniscience | – | – | – | – | – | – | – | 28.8#20 | Artificial Analysis | |
| MMLU-Pro | – | – | – | – | – | – | – | 89.6%#9 | Vals AI | |
| LegalBench | – | – | – | – | – | – | – | 83.6%#48 | Vals AI | |
| CorpFin | – | – | – | – | – | – | – | 66.7%#17 | Vals AI | |
| TaxEval | – | – | – | – | – | – | – | 75.6%#14 | Vals AI |
Instruction following
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| LiveBench Languagenot in index | – | – | – | – | – | 79.7%#26 | – | – | LiveBench | |
| IFBench | – | – | – | – | – | – | – | 62.2%#116 | Artificial Analysis |
Human preference
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| LMArena Text | – | – | – | 1482#18 | – | – | – | 1473#33 | LMArena | |
| EQ-Bench 4 | – | – | – | – | – | – | – | 1281#6 | EQ-Bench |
Multimodal
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| LMArena Vision | – | – | – | 1294#19 | – | – | – | 1289#23 | LMArena |
Long context
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| AA-LCR | – | – | – | – | – | – | – | 77.7%#87 | Artificial Analysis |
Composite
| Benchmark | no reasoning | low | medium | high | xhigh | max | thinking | default | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | – | – | – | – | – | – | 158.3#10 | Epoch AI Benchmarking Hub | |
| LiveBench | – | – | – | – | – | 76.2%#21 | – | – | LiveBench | |
| AA Intelligence Index | – | – | – | – | – | – | – | 42.0#29 | Artificial Analysis | |
| Vals Indexnot in index | – | – | – | – | – | – | – | 60.9#8 | Vals AI |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Anthropic Fast | 126 tok/s | 1.63 s | $10.00 | $50.00 | 1M | – |
| Amazon Bedrock (EU) | 78 tok/s | 1.31 s | $5.50 | $27.50 | 1M | – |
| Amazon Bedrock | 68 tok/s | 2.38 s | $5.00 | $25.00 | 1M | – |
| Claude Platform on AWS | 63 tok/s | 1.73 s | $5.00 | $25.00 | 1M | – |
| Google Vertex | 63 tok/s | 2.78 s | $5.00 | $25.00 | 1M | – |
| Anthropic | 62 tok/s | 1.43 s | $5.00 | $25.00 | 1M | – |
Price history
Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.500 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0095 | $0.0081 | 34.7 s |
| Summarise a 30-page report | 12,000 / 600 | $0.075 | $0.035 | 39.9 s |
| Code edit | 6,000 / 1,500 | $0.068 | $0.047 | 55.5 s |
| Agentic coding session | 60,000 / 4,000 | $0.400 | $0.198 | 1.6 min |
| Structured extraction | 2,000 / 200 | $0.015 | $0.0083 | 33.0 s |
See also
Data as of 9 Sept 2026. Compare these configurations.