Anthropic model product
Claude Opus 4.8
7 and its most capable general-access model at release, with improvements across software engineering, agentic tool use, reasoning, computer use, and knowledge-work benchmarks while shipping at the same price ($5/$25 per million.
Updated Aug 17, 2026. Default version: Claude Opus 4.8
LLMBoard Score87.1Claude Opus 4.8
Coverage60%26 benchmark families
Context window1MTokens
Official input price$5Anthropic API
Claude Opus 4.8 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Claude Opus 4.8 Benchmark Results
Benchmark scores for Claude Opus 4.8.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkInclude | Score87.6% | Rank01 | Participants31 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkScreenSpot Pro | Score87.9% | Rank01 | Participants25 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkDeepSearchQA | Score93.1% | Rank02 | Participants9 | Percentile87.5% | EvidenceC | Evaluated |
| BenchmarkGraphwalks parents >128k | Score83.3% | Rank02 | Participants7 | Percentile83.3% | EvidenceC | Evaluated |
| BenchmarkSWE-bench Multilingual | Score84.4% | Rank02 | Participants38 | Percentile97.3% | EvidenceC | Evaluated |
| BenchmarkOfficeQA Pro | Score66.2% | Rank03 | Participants7 | Percentile66.7% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Multimodal | Score38.4% | Rank03 | Participants4 | Percentile33.3% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Pro | Score69.2% | Rank03 | Participants50 | Percentile95.9% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score88.6% | Rank03 | Participants111 | Percentile98.2% | EvidenceC | Evaluated |
| BenchmarkFinance Agent v2 | Score53.9% | Rank04 | Participants26 | Percentile88.0% | EvidenceB | Evaluated |
| BenchmarkFrontierCode 1.1 | Score46.5% | Rank04 | Participants16 | Percentile80.0% | EvidenceB | Evaluated |
| BenchmarkFrontierSWE | Score75.0% | Rank04 | Participants16 | Percentile80.0% | EvidenceB | Evaluated |
| BenchmarkOSWorld-Verified | Score83.4% | Rank04 | Participants24 | Percentile87.0% | EvidenceC | Evaluated |
| BenchmarkCharXiv-R | Score89.9% | Rank05 | Participants51 | Percentile92.0% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score93.6% | Rank05 | Participants239 | Percentile98.3% | EvidenceC | Evaluated |
| BenchmarkGraphwalks BFS >128k | Score68.1% | Rank05 | Participants11 | Percentile60.0% | EvidenceC | Evaluated |
| BenchmarkMCP Atlas | Score82.2% | Rank05 | Participants33 | Percentile87.5% | EvidenceC | Evaluated |
| BenchmarkCyberGym | Score78.8% | Rank06 | Participants13 | Percentile58.3% | EvidenceC | Evaluated |
| BenchmarkHealthBench Professional | Score55.8% | Rank06 | Participants9 | Percentile37.5% | EvidenceC | Evaluated |
| BenchmarkLiveBench | Score77.2% | Rank06 | Participants38 | Percentile86.5% | EvidenceB | Evaluated |
| BenchmarkFinance Agent | Score53.9% | Rank07 | Participants8 | Percentile14.3% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.0 | Score74.6% | Rank07 | Participants51 | Percentile88.0% | EvidenceC | Evaluated |
| BenchmarkToolathlon | Score59.9% | Rank07 | Participants37 | Percentile83.3% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam | Score57.9% | Rank08 | Participants99 | Percentile92.9% | EvidenceC | Evaluated |
| BenchmarkDeepSWE 1.1 | Score59.0% | Rank12 | Participants25 | Percentile54.2% | EvidenceB | Evaluated |
| BenchmarkBrowseComp | Score84.3% | Rank13 | Participants62 | Percentile80.3% | EvidenceC | Evaluated |
Claude Opus 4.8 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenadocument style control | Categoryoverall | Rank07 | Rating / score1486.6 | Votes8,388 | ObservationsN/A | Result date |
| Arenaagent steerability | Categoryoverall | Rank08 | Rating / score0.1 | VotesN/A | Observations22.4K | Result date |
| Arenaagent praise complaint | Categoryoverall | Rank09 | Rating / score0.1 | VotesN/A | Observations6.3K | Result date |
| Arenaagent task outcome explicit | Categoryoverall | Rank09 | Rating / score0.1 | VotesN/A | Observations20.7K | Result date |
| Arenadocument style control | Categoryoverall | Rank10 | Rating / score1481.8 | Votes8,193 | ObservationsN/A | Result date |
| Arenasearch factuality | Categoryoverall | Rank10 | Rating / score1199.3 | Votes48,610 | ObservationsN/A | Result date |
| Arenasearch style control | Categoryoverall | Rank10 | Rating / score1195.6 | Votes48,610 | ObservationsN/A | Result date |
| Arenasearch | Categoryoverall | Rank11 | Rating / score1204.9 | Votes48,610 | ObservationsN/A | Result date |
| Arenadocument | Categoryoverall | Rank12 | Rating / score1474.6 | Votes8,388 | ObservationsN/A | Result date |
| Arenaagent bash recovery steps | Categoryoverall | Rank13 | Rating / score0.1 | VotesN/A | Observations24.2K | Result date |
| Arenavision style control | Categoryoverall | Rank14 | Rating / score1284.1 | Votes11,131 | ObservationsN/A | Result date |
| Arenadocument | Categoryoverall | Rank16 | Rating / score1468.9 | Votes8,193 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank18 | Rating / score1539.0 | Votes9,906 | ObservationsN/A | Result date |
| Arenavision | Categoryoverall | Rank21 | Rating / score1292.5 | Votes11,131 | ObservationsN/A | Result date |
| Arenavision style control | Categoryoverall | Rank21 | Rating / score1280.4 | Votes11,571 | ObservationsN/A | Result date |
| Arenavision | Categoryoverall | Rank22 | Rating / score1289.7 | Votes11,571 | ObservationsN/A | Result date |
| Arenaagent | Categoryoverall | Rank23 | Rating / score0.0 | VotesN/A | Observations1.4M | Result date |
| Arenatext style control | Categoryoverall | Rank26 | Rating / score1473.6 | Votes41,219 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank30 | Rating / score1463.7 | Votes41,146 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank34 | Rating / score1452.5 | Votes41,219 | ObservationsN/A | Result date |
| Arenaagent tool hallucination | Categoryoverall | Rank48 | Rating / score-0.3 | VotesN/A | Observations1.3M | Result date |
Claude Opus 4.8 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $5 input, $25 output per 1M
- Official provider
- Anthropic
- Lowest third-party
- From $0.425 input, $2.13 output per 1M via UnoRouter
- Tracked offerings
- 41
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderKenari | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderUnoRouter | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$0.425 | Output / 1M$2.13 | Context1M | Updated |
| ProviderXpersona | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$1.5 | Output / 1M$9.25 | Context200K | Updated |
| ProviderPoe | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$4.29 | Output / 1M$21.46 | Context1M | Updated |
| ProviderNanoGPT | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderVertex | Provider model IDclaude-opus-4-8@default | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderImpossibl | Provider model IDanthropic/claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderOpenRouter | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderVertex (Anthropic) | Provider model IDclaude-opus-4-8@default | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderDaoXE | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderCrossModel | Provider model IDanthropic/claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAnthropic | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| Providerrouting.run | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAmazon Bedrock | Provider model IDanthropic.claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderMerge Gateway | Provider model IDanthropic/claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderCloudflare AI Gateway | Provider model IDanthropic/claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderFastRouter | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderZenMux | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderGMI Cloud | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderFreeModel | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderOfox | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderNeon | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAzure Cognitive Services | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderOpenCode Zen | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderRequesty | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAzure | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAIHubMix | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context200K | Updated |
| ProviderAIHubMix | Provider model IDclaude-opus-4-8-think | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context200K | Updated |
| ProviderVercel AI Gateway | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderEden AI | Provider model IDanthropic/claude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderLLM Gateway | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderGitHub Copilot | Provider model IDclaude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context200K | Updated |
| ProviderModelis | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderKilo Gateway | Provider model IDanthropic/claude-opus-4.8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderPioneer | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderAbacus | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$5 | Output / 1M$25 | Context1M | Updated |
| ProviderCortecs | Provider model IDclaude-opus4-8 | Regionglobal | Input / 1M$5.44 | Output / 1M$27.19 | Context1M | Updated |
| ProviderVenice AI | Provider model IDclaude-opus-4-8 | Regionglobal | Input / 1M$6 | Output / 1M$30 | Context1M | Updated |
| ProviderKilo Gateway | Provider model IDanthropic/claude-opus-4.8-fast | Regionglobal | Input / 1M$10 | Output / 1M$50 | Context1M | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Claude Opus 4.8 Runtime Performance
Provider-specific output speed and catalog latency for Claude Opus 4.8. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderAnthropic | Output Speed45.954 tok/s | Catalog Latency1.29 s | Max Input1M | Max Output128K | Updated |
| ProviderVertex AI | Output Speed42 tok/s | Catalog Latency0.5 s | Max Input1M | Max Output128K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Claude Opus 4.8 Specifications
Technical details for the model's default version.
- Version
- Claude Opus 4.8
- Released
- May 28, 2026
- Knowledge cutoff
- Jan 1, 2026
- Parameters
- N/A
- Context window
- 1M
- Max output
- 128K
- Inputs
- image, pdf, text
- Outputs
- text
- Open weights
- No
- License
- Proprietary
Claude Opus 4.8 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionClaude Opus 4.8 | Released | LLMBoard87.1 | ParametersN/A | Context1M | Max output128K | Open weightsNo | LicenseProprietary |
Claude Opus 4.8 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Claude Opus 4.8
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AN
Claude Opus 4.7
Anthropic
AN
Claude Sonnet 5
Anthropic
AN
Claude Mythos
Anthropic
AN
Claude Opus 4.6
Anthropic
AN
Claude Fable 5
Anthropic
AN
Claude Opus 5
Anthropic
What is Claude Opus 4.8?
Key information about Claude Opus 4.8 and its available data.
7 and its most capable general-access model at release, with improvements across software engineering, agentic tool use, reasoning, computer use, and knowledge-work benchmarks while shipping at the same price ($5/$25 per million input/output tokens). 2%), and GDPval-AA (1890 Elo).
6 on dishonest agentic code summaries, and broadly improved adherence to Claude's constitution. The model defaults to high effort and exposes new 'extra' (xhigh) and 'max' levels for harder problems.
ai and Cowork, and a Messages API extension that accepts system entries inside the messages array so harnesses can update instructions mid-task without breaking the prompt cache. 5× speed at $10/$50 per million input/output tokens, three times cheaper than fast mode on previous models.
Available across Claude products, the Claude API as `claude-opus-4-8`, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
Data as of 2026-08-17.
FAQ
Common questions about Claude Opus 4.8.
When was Claude Opus 4.8 released?
Claude Opus 4.8's default version was released on May 28, 2026.
How much does Claude Opus 4.8 cost?
Claude Opus 4.8's official API price is $5 per million input tokens and $25 per million output tokens via Anthropic. The lowest tracked third-party offer starts at $0.425 input and $2.13 output via UnoRouter.
Who created Claude Opus 4.8?
Claude Opus 4.8 was created by Anthropic.
What is the context window for Claude Opus 4.8?
The default version has a 1M token context window.
Is Claude Opus 4.8 open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer Claude Opus 4.8?
41 provider offerings are linked to the default version.
What models should I compare Claude Opus 4.8 with?
Nearby ranked alternatives include Qwen3.8 Max, Muse Spark 1.1, GPT-5.6-Terra.