text model
Claude Opus 5
AnthropicLong-horizon reasoning and the steadiest instruction-following in the catalogue.
Opus 5 is the model to reach for when a task has many steps and the cost of drifting off-instruction is high. In hiring that maps almost exactly onto rubric-based scoring: it holds a long rubric and applies it consistently across a batch, which is the property that makes scores defensible.
CapabilitiesExtended thinkingTool useVisionStructured outputsPrompt caching
Variants
3 ways to run it.
Same family, different quality/cost rung. Picking the wrong one is the most common way teams overpay.
Opus 5
Maxclaude-opus-5Frontier tier for the hardest work.
15×
rel. cost
96
quality
Sonnet 5
Balancedclaude-sonnet-5The production workhorse.
3×
rel. cost
89
quality
Haiku 4.5
Fastclaude-haiku-4-5Fast and cheap for high-volume steps.
1×
rel. cost
76
quality
Fit
Two audiences, one model.
The same model is a different proposition depending on what you point it at.
91/100
Recruitment fit
The best choice for rubric-based scoring, where consistency across a batch matters more than peak score.
Best for
- Structured scoring
- Bias-audit passes over model output
- Candidate feedback drafting
Strengths
- Applies a long rubric consistently across many candidates
- Strong at citing which part of a CV drove a judgement — useful for audit trails
Watch out for
- Opus pricing makes it unsuitable for first-pass parsing
- Use Haiku for the cheap steps
