Claude Opus 4.8 Thinking
Anthropic Claude Opus 4.8 with thinking enabled.
- Reasoning
- Vision
- Native PDF input
- Tool Calling
- Structured Output
Added May 28, 2026
Pricing
Auto routing · per 1M tokens- Input
- $5.00
- Output
- $25.00
- Cache read
- $0.50
Specifications
- Context window
- 1M
- Max output
- 128K
- Avg output (7d)
- 974 tokens
- Longer than 70% of models
Benchmarks
Benchmarks
Sourced from Artificial Analysis.
Intelligence Index
41.8
Coding Index
74.3
Agentic Index
41.9
Agentic work
AutomationBench-AA
Workflow automation with guardrail penalties
45.6%
Better than 60% of models compared
Harvey LAB-AA
Legal agentic work criterion pass rate
91.1%
Better than 65% of models compared
AA-Briefcase
Agentic knowledge work (Elo)
1321 Elo
Better than 71% of models compared
GDPval-AA v2
Economically valuable tasks (Elo)
1456 Elo
Better than 81% of models compared
Document reasoning
GDP.pdf
Professional PDF reasoning: all-pass rate
22.8%
Better than 79% of models compared
AA-LCR v1.1
Long context reasoning with updated grading
77.7%
Better than 78% of models compared
MLCR-AA
Medical long-context reasoning
45.6%
Better than 86% of models compared
Reasoning
HLE
Humanity's Last Exam
48.7%
Better than 95% of models compared
IFBench
Instruction-following benchmark
62.2%
Better than 73% of models compared
CritPt
Research-level physics reasoning
20.9%
Coding
Terminal-Bench v4.0
Practical coding and terminal tasks
21.7%
Better than 75% of models compared
SciCode
Python programming for scientific computing
54.4%
Better than 74% of models compared
Knowledge
AA-Omniscience Accuracy
Proportion of correctly answered questions
48.8%
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
39.3%
Legacy benchmarks
GPQA Diamond (legacy)
Graduate-level scientific reasoning
92.0%
Better than 93% of models compared
Terminal-Bench Hard (legacy)
Agentic coding and terminal use
58.3%
Better than 98% of models compared
T²-Bench Telecom (legacy)
Conversational AI agents in dual-control scenarios
94.4%
Better than 93% of models compared
AA-LCR (unversioned / legacy)
Long context reasoning evaluation
77.7%
Better than 78% of models compared
GDPval-AA (unversioned / legacy)
Economically valuable tasks
47.7%
Last updated Oct 14, 2026
Artificial AnalysisProviders
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…