Claude Opus 4.8
anthropic/claude-opus-4-8
Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents
Capabilities
ReasoningYes
Tool callingYes
Structured outputNo
TemperatureFixed
AttachmentYes
Open weightsNo
Limits
Context1,000,000 tokens
Max inputN/A
Max output128,000 tokens
Details
Familyclaude-opus
Knowledge cutoff2026-01
Release date2026-05-28
Last updated2026-05-28
Modalities
Inputtext, image, pdf
Outputtext
Benchmarks
SWE-Bench Pro
69.2
resolve rate
Terminal-Bench
74.6
success rate
SWE-Bench Verified
88.6
resolved
Humanity's Last Exam
49.8
accuracy
Humanity's Last Exam
57.9
accuracy
OSWorld-Verified
83.4
success rate
FrontierCode
13.4
pass rate
Community Rating
—
No ratings yet
Sign in to rate
Pricing
Cost per 1M tokens
InputN/A
OutputN/A