Constitutional AI
Harmlessness from AI Feedback
Anthropic's alternative to RLHF: instead of using human raters to rank responses, the model critiques and revises its own outputs according to a written list of principles (a "constitution"). This reduces the need for expensive human labeling and makes the alignment process more transparent — anyone can read the constitution.
Why it matters
Whether you trust Anthropic's safety story or not, this paper reshaped the alignment vocabulary. Every frontier lab now publishes some version of "model principles," "spec," or "policy" that traces back to the constitutional framing introduced here.
Core Capabilities
Long Documents
Handles entire codebases, books, and multi-doc RAG.
Research
Foundational paper or scientific contribution.
Context Window
Context window not disclosed.
Availability
API
Not available
Product / App
Not available
Open Source
Not released
Enterprise
—
Pricing Model
Research artifact
Not commercially released.
Research Capability / Performance
Where this model sits relative to the middle 60% of models in the tree. All scores are 0–10 (higher is better).
Lower 20% Upper 80% This model
Lower 20% 20th percentile — 20% of models score below this This model Where the current model lands Upper 80% 80th percentile — only 20% of models score above this
Percentile boundaries are computed across every model in the tree that reports the underlying benchmark for each capability.
What it feels like
- Language model from Anthropic — see the linked sources below for benchmark and review coverage
Best use cases
- General-purpose tasks within Anthropic's deployment footprint
- See the model spec and sources block for benchmarked use cases
Tools to try
Not ideal for
- Tasks far outside the modalities listed in this model's spec
- Workflows where a more recent successor in the same family scores higher