Milvus vs Weaviate
Milvus vs Weaviate: side-by-side comparison of two vector-database infrastructure systems — architecture, taxonomy, license, pricing, MCP/A2A support, and direct edges.
Cost & capability
| Milvus | Weaviate | |
|---|---|---|
| Cost tier | free | free |
| $/Mtok input | 0 | 0 |
| $/Mtok output | 0 | 0 |
Where they differ (12)
Rows where both sides have data and the values disagree — the shortlist of dimensions that actually distinguish these two systems.
| Milvus | Weaviate | |
|---|---|---|
| Type | Multi-vector + hybrid + GPU | Hybrid vector DB + learned ranking |
| Created | 2019-09 | 2016-03 |
| Latest release | v3.0.3 (2026-10-14) | v1.40.2 (2026-10-21) |
| License | Apache-2.0 | BSD-3-Clause |
| GitHub | milvus-io/milvus — ~46k stars, last commit 2026-10 | weaviate/weaviate — ~16.8k stars, last commit 2026-10 |
| Pricing | free (OSS); managed via Zilliz Cloud (free tier, then usage-based) | Free cluster $0 (100k objects, 1 GB RAM, 10 GB disk); Flex pay-as-you-go from $45/mo; Premium prepaid via sales; OSS free |
| Funding | $113M total (Zilliz: $43M 2020 + $60M 2022); Zilliz is the commercial entity | $68M total $200M val Series B · 2023-04 |
| Backend storage | custom (object-storage backed; etcd metadata; Pulsar/Kafka log) | custom (LSM-backed vector + inverted index) |
| Deployment | Both (self-hosted OSS + Zilliz Cloud managed) | Both (self-hosted + managed cloud) — OSS (Docker/K8s); Weaviate Cloud shared (Free/Flex) and dedicated (Premium) |
| API surface | REST, gRPC, SDK: Python, Java, Go, Node.js, C#, Ruby | REST, GraphQL, gRPC, SDK: Python, JS/TS, Go, Java |
| Multi-tenancy | Control plane and data plane in separate VPCs/subnets; BYOC option for hard tenant isolation | Tenant-per-collection (multi-tenancy feature) gives logical isolation without separate clusters; Dedicated Cloud option for hard isolation |
| MCP | yes — official zilliztech/mcp-server-milvus (GitHub) | yes — official MCP server built into Weaviate since v1.38 (Streamable HTTP at /v1/mcp) |
At a glance
| Milvus | Weaviate | |
|---|---|---|
| Section | Vector-database infrastructure | Vector-database infrastructure |
| Tier | T1 | T1 |
| Type | Multi-vector + hybrid + GPU | Hybrid vector DB + learned ranking |
| Created | 2019-09 | 2016-03 |
| Latest release | v3.0.3 (2026-10-14) | v1.40.2 (2026-10-21) |
| License | Apache-2.0 | BSD-3-Clause |
| GitHub | milvus-io/milvus — ~46k stars, last commit 2026-10 | weaviate/weaviate — ~16.8k stars, last commit 2026-10 |
| Pricing | free (OSS); managed via Zilliz Cloud (free tier, then usage-based) | Free cluster $0 (100k objects, 1 GB RAM, 10 GB disk); Flex pay-as-you-go from $45/mo; Premium prepaid via sales; OSS free |
| Funding | $113M total (Zilliz: $43M 2020 + $60M 2022); Zilliz is the commercial entity | $68M total $200M val Series B · 2023-04 |
| Backend storage | custom (object-storage backed; etcd metadata; Pulsar/Kafka log) | custom (LSM-backed vector + inverted index) |
| Deployment | Both (self-hosted OSS + Zilliz Cloud managed) | Both (self-hosted + managed cloud) — OSS (Docker/K8s); Weaviate Cloud shared (Free/Flex) and dedicated (Premium) |
| API surface | REST, gRPC, SDK: Python, Java, Go, Node.js, C#, Ruby | REST, GraphQL, gRPC, SDK: Python, JS/TS, Go, Java |
| Embedding | multiple supported | multiple supported |
| Multi-tenancy | Control plane and data plane in separate VPCs/subnets; BYOC option for hard tenant isolation | Tenant-per-collection (multi-tenancy feature) gives logical isolation without separate clusters; Dedicated Cloud option for hard isolation |
| MCP | yes — official zilliztech/mcp-server-milvus (GitHub) | yes — official MCP server built into Weaviate since v1.38 (Streamable HTTP at /v1/mcp) |
| A2A | no Google A2A (Agent2Agent) integration documented as of 2026-05. | no Google A2A (Agent2Agent) integration documented as of 2026-05. |
| OpenTelemetry | first-class — Prometheus + OTel | first-class — Prometheus + OTel |
| Optimised for | low-latency similarity search + scale | low-latency similarity search + scale |
| Anti-fit | not for relational / graph-heavy queries (vector-first by design) | not for relational / graph-heavy queries (vector-first by design) |
Pros & cons
Milvus
Pros: Largest-scale OSS vector DB — billions of vectors comfortably; CNCF graduate; strong community in China + research labs.
Cons: Operational complexity is high vs Pinecone-managed; documentation can be uneven across language clients.
Weaviate
Pros: OSS-first managed and self-hosted vector DB with native hybrid (vector + keyword) and module ecosystem (Q&A, summarization, classification).
Cons: Module surface area can lead to feature drift; smaller enterprise sales presence than Pinecone.