Back to Arena

Abridge

by Abridge

System Card

OrganizationAbridge
Released2018-01
Architectureagentic-workflow / Contextual Reasoning Engine for Clinical Documentation
DetailsAbridge developed its own proprietary speech recognition and note generation models, eliminating reliance on third-party LLMs, with a Contextual Reasoning Engine that links the current clinical conversation to prior visit context, specialty-specific workflows, and billing codes. The platform delivers complete draft notes within under 2 minutes after a visit. Used in 3M+ patient visits across 600+ health organizations; raised $150M Series C. Architecture provides citation-level traceability from note to source conversation.
Parameters—
Domainepisodic-session
Open SourceNo
WebsiteVisit
healthcareambient-scribeclinical-notesproprietary-llmepic-integrationbilling-codes

Capability Profile

Benchmark Scores

4 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
2/2
LoCoMo
79.482p
MemoryBank
77.460p
Cross-Session Memory
1/1
LongMemEval
75.146p
Multi-Hop QA
0/3
BABILong
no data
MultiHop-RAG
no data
HotpotQA
no data
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Abridge vendor documentation; evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Abridge vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Abridge vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Abridge vendor documentation; evaluated on MemoryBank: Enhancing LLMs with Long-Term Memory (Sun Yat-sen University, 2305)