Context window and long-document fit
Review this signal against your actual prompt distribution and provider documentation.
Builder decision page
Claude Fable 5 for RAG: Context, Retrieval, and Cost
Compare retrieval-augmented generation fit using context capacity, pricing, operational trade-offs, and current source-grounded evidence. Values below come from the canonical AIBuzzHub model profile and are marked as not verified when unavailable.
Builder answer
RAG decisions should be made from context capacity, input economics, retrieval quality, and production reliability—not headline benchmark scores alone.
Input / 1M tokens
$10.00
Output / 1M tokens
$50.00
Context window
—
Weights
API / closed
for RAG signals
Input-token economics for retrieved passages
Review this signal against your actual prompt distribution and provider documentation.
Operational trade-offs for grounded answers
Review this signal against your actual prompt distribution and provider documentation.
Operational and benchmark data
Latency p50
Not verified
TTFT
Not verified
Throughput
Not verified
License
Proprietary API
Operational fields and scores are only decision aids. Confirm methodology, model version, region, and workload-specific behavior before shipping.
Builder note
Standard API list pricing. Cache reads, cache writes, Batch API, Fast mode, and US-only inference can change the effective cost.