anthropic

Claude Opus 4.6

claude-opus

High-end Claude for difficult coding, planning, and slower expert reasoning

ReasoningTool CallingAttachments

Context Window

1M

Max Output

128K

Temperature

Yes

Open Weights

No

Knowledge Cutoff

2025-05-31

Released

2026-02-05

Last Updated

2026-03-13

Modalities

Input:TextImagePDF
→
Output:Text

Benchmarks

BenchmarkScoreMetricHarness
SWE-Bench Pro51.9resolve rate-
SWE-Atlas Codebase QnA33.3scoreClaude Code
SWE-Atlas Codebase QnA30scoreMini-SWE-Agent
SWE-Atlas Refactoring35.58scoreClaude Code
SWE-Atlas Test Writing36.67scoreClaude Code
SWE-Atlas Test Writing36.08scoreMini-SWE-Agent
Artificial Analysis Coding Agent Index51.3average pass@1Claude Code
SWE-Atlas Codebase QnA71.9pass@1Claude Code
SWE-Bench Pro11.8pass@1Claude Code
Terminal-Bench70.2pass@1Claude Code

Available Providers (6)

ProviderInput /1MOutput /1MCache Read /1MCache Write /1MReasoningStatus
RequestyRequestyRp 89.482Rp 447.408Rp 8.949Rp 111.852effort, budget_tokens—
AnyAPIAnyAPI————toggle, effort, budget_tokensExperimental
Merge GatewayMerge GatewayRp 89.482Rp 447.408Rp 8.949Rp 111.852toggle, effort, budget_tokensExperimental
NEAR AI CloudNEAR AI CloudRp 89.482Rp 447.408Rp 8.949Rp 111.852toggle, budget_tokens—
Cloudflare AI GatewayCloudflare AI GatewayRp 89.482Rp 447.408Rp 8.949Rp 111.852effort—
Perplexity AgentPerplexity AgentRp 89.482Rp 447.408Rp 8.949—effort—
← Kembali ke Katalog