Anthropic: Claude Opus 4.7, Anthropic — route, pricing & source
Anthropic generally available frontier model for complex reasoning, coding, vision, and long-horizon agentic work.
Context and limits
From source metadata + verification fields — confirm provider docs before production.
- LMArena Text Arenaoverall · official_benchmark_api1,503Independent
Rank: 1 | Votes: 8945 | 95% CI: 1496.6-1509.4 | Published: 2026-05-07 | LMArena dataset entry: "claude-opus-4-7-thinking" (matched via variant normalization) | Model license label in dataset: Proprietary
- LMArena WebDev Arenaoverall · official_benchmark_api1,570Independent
Rank: 1 | Votes: 3309 | 95% CI: 1558.7-1582.2 | Published: 2026-05-07 | LMArena dataset entry: "claude-opus-4-7-thinking" (matched via variant normalization) | Model license label in dataset: Proprietary
- LMArena Vision Arenaoverall · official_benchmark_api1,305Independent
Rank: 1 | Votes: 4666 | 95% CI: 1295.5-1315.3 | Published: 2026-05-07 | LMArena dataset entry: "claude-opus-4-7-thinking" (matched via variant normalization) | Model license label in dataset: Proprietary
Access and license
Example costs
Scenario: 1K input + 300 output tokens / request
Cheaper alternatives
Lower-cost models for the same scenario
Source
- Official Anthropic docs confirm Opus 4.7, 1M context, 128K max output, and public pricing.
- No independent benchmark score has been merged yet for this model.
- Includes Hugging Face benchmark data only when the dataset is public, ungated, allowlisted, and matched by reviewed aliases.
