Chroma vs pgvector
Chroma vs pgvector: side-by-side comparison of two vector-database infrastructure systems — architecture, taxonomy, license, pricing, MCP/A2A support, and direct edges.
Cost & capability
| Chroma | pgvector | |
|---|---|---|
| Cost tier | free | free |
| $/Mtok input | 0 | 0 |
| $/Mtok output | 0 | 0 |
Where they differ (15)
Rows where both sides have data and the values disagree — the shortlist of dimensions that actually distinguish these two systems.
| Chroma | pgvector | |
|---|---|---|
| Type | Embedded vector DB (Rust core) | PostgreSQL vector extension |
| Created | 2022-10 | 2021-04 |
| Latest release | 1.5.9 2026-05-05 | no releases |
| License | Apache-2.0 | Custom |
| GitHub | 27.8k★ +125/mo Rust | 21.1k★ +123/mo C |
| Pricing | Free + paid | Free (PostgreSQL License open source) |
| Funding | $18M total $75M val Seed · 2023-04 | No external funding — open-source project by independent developer |
| Backend storage | SQLite (local) / custom cloud backend | Postgres |
| Deployment | Both | Self-hosted only (PostgreSQL extension; managed via Supabase/AWS RDS/etc.) |
| API surface | REST, SDK: Python, JS/TS | SQL via Postgres protocol |
| Embedding | multiple supported | BYO |
| Multi-tenancy | Multi-tenant indexes with billions of vectors; AWS PrivateLink; BYOC (Bring Your Own Cloud) for hard isolation | Logical isolation via Postgres schemas/databases; managed providers (Supabase, Neon, RDS) add tenant-level controls |
| MCP | via official adapter — chroma-mcp | via Postgres MCP servers (community) |
| A2A | no Google A2A (Agent2Agent) integration documented as of 2026-05. | not supported |
| OpenTelemetry | via OpenTelemetry instrumentation | via Postgres OTel |
At a glance
| Chroma | pgvector | |
|---|---|---|
| Section | Vector-database infrastructure | Vector-database infrastructure |
| Tier | T1 | T1 |
| Type | Embedded vector DB (Rust core) | PostgreSQL vector extension |
| Created | 2022-10 | 2021-04 |
| Latest release | 1.5.9 2026-05-05 | no releases |
| License | Apache-2.0 | Custom |
| GitHub | 27.8k★ +125/mo Rust | 21.1k★ +123/mo C |
| Pricing | Free + paid | Free (PostgreSQL License open source) |
| Funding | $18M total $75M val Seed · 2023-04 | No external funding — open-source project by independent developer |
| Backend storage | SQLite (local) / custom cloud backend | Postgres |
| Deployment | Both | Self-hosted only (PostgreSQL extension; managed via Supabase/AWS RDS/etc.) |
| API surface | REST, SDK: Python, JS/TS | SQL via Postgres protocol |
| Embedding | multiple supported | BYO |
| Multi-tenancy | Multi-tenant indexes with billions of vectors; AWS PrivateLink; BYOC (Bring Your Own Cloud) for hard isolation | Logical isolation via Postgres schemas/databases; managed providers (Supabase, Neon, RDS) add tenant-level controls |
| MCP | via official adapter — chroma-mcp | via Postgres MCP servers (community) |
| A2A | no Google A2A (Agent2Agent) integration documented as of 2026-05. | not supported |
| OpenTelemetry | via OpenTelemetry instrumentation | via Postgres 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) |
Taxonomy
| Axis | Chroma | pgvector |
|---|---|---|
| storage | vector | vector |
| retrieval | similarity | similarity |
| persistence | long-term | long-term |
| update | overwrite | overwrite |
| unit | chunk | chunk |
| governance | inspectable | inspectable |
| conflict | overwrite | overwrite |
Pros & cons
Chroma
Pros: Lowest-friction local vector store — embedded mode means zero ops for prototyping; popular default in agent tutorials.
Cons: Production scale-out story is less mature than Weaviate / Qdrant; managed cloud (Chroma Cloud) is newer.
pgvector
Pros: Postgres extension — keeps vector data alongside transactional data with full ACID; minimal ops for shops already running Postgres.
Cons: Performance plateau at very large vector counts; limited filter pushdown vs purpose-built engines.