Back to Arena

Vellum AI

by Vellum AI Inc. (YC W23)

System Card

OrganizationVellum AI Inc. (YC W23)
Released2023-02
Architectureagentic-workflow / LLM workflow and evaluation platform
DetailsVellum provides an enterprise platform for building, evaluating, and deploying LLM workflows (including RAG pipelines). Features include prompt versioning, A/B testing, workflow orchestration with a visual editor, regression testing suites, and production monitoring with human-in-the-loop feedback. Raised $25.5M total ($5M seed, $20M Series A). YC W23 company.
Parameters—
Domainrag-retrievalagent-memory
Open SourceNo
WebsiteVisit
prompt-managementworkflow-orchestrationevaluationA/B-testingenterprise

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
1/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
603p
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
74.643p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
77.460p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
7579p
HotpotQA
75.170p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Vellum AI vendor documentation; evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Vellum AI vendor documentation; evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)Vellum AI vendor documentation; evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Vellum AI vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Vellum AI vendor documentation; evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)Vellum AI vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)