Pick this when
The last few points of capability when you have already measured an edge over Opus.
Skip when
You need zero data retention, or you picked it only because the catalog lists it above Opus.
Nominal flagship. Independent scores sit one point behind Opus 5 at twice the price. Treat the name as marketing until your own eval says otherwise.
Price
- Input / 1M
- $10
- Output / 1M
- $50
- Measured task
- $3.14
Covered Model: 30-day retention, no zero-data-retention. Safety routing can fall back to Opus 4.8.
Shape
- Context
- 1M
- Max output
- 128K
- Released
- 2026-07
- Tools
- Yes
- Reasoning
- Yes
Scores
Intelligence index62
GPQA Diamond92
Terminal-Bench 2.185
Agentic index57
- Arena
- 2003
- Speed
- 106/s
- GPQA
- 91.8%
- Term.
- 84.6%
- AA Index. The single number most people mean by “how smart.” It is a blend, so a specialist can lose here and still win the job you care about.
- GPQA. A clean test of hard reasoning. PhD experts sit around 65%. It says little about writing, tools, or taste.
- Term.. Closest public proxy for coding agents that live in a terminal. A high index score with a weak terminal score is a warning.
- Arena. Captures taste and usefulness that unit tests miss. Sample size varies by model, so treat gaps under ~100 Elo as noise.
Anthropic · San Francisco. Claude. Opus 5 leads public intelligence rankings. Fable 5 is the named flagship and the expensive one. Sonnet 5 is the daily driver.