Weaviate
Open-core: free OSS vector database monetized via managed cloud (serverless and dedicated Weaviate Cloud), enterprise support/licensing, and cloud-marketplace consumption (AWS/GCP/Azure)
Only the Series B mark (~$200M) is externally reported, and by press (The Information), not the company. Seed and Series A post-money figures were never disclosed; the plotted values are flagged inferences from round size and typical dilution, included only to shape the trail. Aggregator listings of a '$50M Series C at $200M, Oct 2025' are excluded as a mislabeled duplicate of the Series B (absent from Tracxn's funding-rounds page, Weaviate's investor page, and all tier-1 coverage). The Jun 2026 Ricoh strategic investment discloses no valuation; the trail carries the Series B mark flat.
Earnings, margins, COGS & capex
Weaviate is private and discloses no financials. Verified funding: $67.7M disclosed total across a $1.6M seed (Aug 2020, led by Zetta Venture Partners with ING Ventures, raised as SeMI Technologies), $16M Series A (Feb 2022, co-led by NEA and Cortical Ventures), and $50M Series B (Apr 2023, led by Index Ventures with Battery Ventures, plus NEA, Cortical, Zetta, ING Ventures) at a ~$200M valuation reported by The Information. In June 2026 Ricoh's corporate venture fund (RICOH Innovation Fund) made a strategic investment of undisclosed size. Getlatka estimates ~$12.3M ARR for 2024 - unverified. Headcount estimates conflict: ~104 (Getlatka, 2024), 80+ (Tracxn), ~74 (one tracker, Apr 2026). No confirmed new priced round in the ~3 years since the Series B, implying either disciplined burn on the $50M or a challenging up-round environment as vector search commoditizes - while closest OSS rival Qdrant closed a $50M Series B in Mar 2026.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~40¢ is cost of goods and ~60¢ operating expense, leaving ~0¢ of operating profit.
Revenue trend
Margins
Cloud/serverless mix vs self-hosted OSS determines margin; serverless tier (2024) aims at better unit economics
Presumed negative; typical for a Series B infra company scaling GTM
COGS structure
Primarily hyperscaler compute/storage for Weaviate Cloud (memory-heavy HNSW indexes make RAM the dominant cost driver; compression/off-loading features exist to cut it), plus support engineering. OSS self-hosted users carry their own infra cost.
Capex
Minimal owned capex - no datacenters; capacity rented from AWS/GCP/Azure. Spend is opex-shaped (cloud bills, R&D headcount).
Latest earnings
n/a
n/a - none issued
- Last priced round
- $50M Series B, Apr 2023, led by Index Ventures
- Reported valuation
- ~$200M (Series B, Apr 2023, per The Information; not stated in the company PR)
- Total disclosed raised
- $67.7M through Series B, plus undisclosed Ricoh strategic investment (Jun 2026)
- OSS downloads
- 20M+ (company-stated, 2026)
- Headcount
- ~74-105 (conflicting tracker estimates, 2024-2026)
Growth drivers
- RAG and agentic AI adoption pulling demand for retrieval infrastructure
- Open-source funnel — 20M+ downloads (company-stated, weaviate.io homepage 2026, up from 2M+ at the Apr 2023 Series B) converting to Weaviate Cloud
- Serverless and dedicated cloud tiers lowering entry cost and expanding usage-based revenue
- Weaviate Agents suite (launched Mar 2025) and native hybrid search (vector + keyword/BM25) plus multi-tenancy targeting enterprise SaaS builders
- Cloud-marketplace distribution (AWS, GCP, Azure) shortening enterprise procurement; Ricoh strategic partnership (2026) as a channel into document/data-capture enterprises
Bull & bear
Weaviate is a leading open-source-first, AI-native vector database with real enterprise logos, riding a structural boom in RAG and agentic workloads, at a stale ~$200M mark that leaves substantial upside if cloud conversion compounds.
- Category tailwind: every production LLM app needs retrieval; vector-database spend is growing and Weaviate is a top-3 independent brand in it
- Open-source distribution is a durable moat vs closed rivals like Pinecone - 20M+ downloads seed a self-serve cloud funnel with near-zero CAC
- Hybrid search, multi-tenancy, and BYOC/on-prem deployment fit regulated enterprises that cannot use hyperscaler-native or closed-SaaS options; Morningstar and Instabase case studies prove the enterprise motion
- Entry price is a 2023 mark (~$200M); comparable AI-infra assets re-rated sharply since - even a modest up-round or strategic acquisition clears it, and Ricoh's 2026 strategic investment signals corporate acquirer interest in the asset class
- Capital-efficient: $67.7M disclosed raised, ~100 people; does not need Pinecone-scale burn to reach the same market
Standalone vector databases are being commoditized from below (pgvector, OSS self-hosting) and absorbed from above (MongoDB, Elastic, hyperscalers, model providers); Weaviate is sub-scale, has not confirmed a priced round in 3+ years, and its open-source core caps paid conversion.
- Vector search became a feature, not a product: Postgres pgvector is free, MongoDB (MDB) and Elastic (ESTC) bundle it into databases enterprises already pay for, and AWS/Azure/GCP ship native options; Elastic's CEO called standalone vector DBs 'never a business' (2025)
- ~$12M estimated ARR (2024) roughly five years after the 2020 seed and on $68M raised implies the OSS-to-paid conversion is hard; the biggest users can self-host forever
- No verified priced round since Apr 2023 during the strongest AI-infra funding market in history is itself a negative signal - closest OSS rival Qdrant raised a $50M Series B in Mar 2026 while Weaviate's only new money is an undisclosed corporate strategic check
- Better-capitalized direct rivals (Pinecone raised $138M total incl. $100M at $750M in 2023; Zilliz $113M; Qdrant now ~$88M) can outspend on enterprise GTM, while cheaper serverless entrants (Turbopuffer) undercut on price
- The best-funded pure-play, Pinecone, reportedly hired bankers to explore a sale in 2025 (The Information) after losing marquee customer Notion - if the category leader cannot stand alone, the independent vector DB TAM may be consolidating at unfavorable prices
What it is worth
Last priced round anchor plus ARR-multiple cross-check against public and private AI-infra comps
~$75M-150M
vector search fully commoditizes, ARR growth stalls, and the company consolidates into a larger data platform below the 2023 mark
~$200M-350M
steady but unspectacular growth; next round modestly above the Series B mark or a fair-value strategic acquisition
~$500M-750M
ARR compounds to $30M+ with strong cloud conversion, agentic-AI expansion, and scarcity value as one of the last independent OSS vector platforms; strategic acquirers (data-platform, hyperscaler, or corporates like Ricoh) pay a premium
Last verified mark is ~$200M (Series B, Apr 2023, per The Information) - roughly 16x the ~$12.3M 2024 estimated ARR at that mark, in line with 2023 AI-infra pricing. The mark is 3 years stale: if ARR has compounded toward $25-40M, private comps (10-20x forward ARR for growing AI infra) support meaningfully more; if growth stalled amid commoditization, a down-round or sub-mark acquisition is realistic. The undisclosed Ricoh strategic investment (Jun 2026) provides no valuation read. Aggregator listings of a 2025 'Series C' are treated as a duplicate of the Series B, not a market-clearing event. Category context cuts both ways: Pinecone (last at $750M) reportedly exploring a sale caps optimism, while Qdrant's Mar 2026 $50M Series B shows investors still funding the category. Not financial advice.
SWOT
Strengths
- Genuinely popular open-source core (BSD-3) with strong developer mindshare and 20M+ downloads - low-cost top-of-funnel
- AI-native feature set — built-in hybrid search, model-provider integrations (OpenAI, Cohere, Anthropic, Hugging Face), multi-tenancy, and Weaviate Agents for RAG/agent tooling
- Credible enterprise references (Morningstar, Stack Overflow, Instabase - documented in weaviate.io case studies)
- Tier-1 investor syndicate (Index, Battery, NEA) plus a corporate strategic backer (Ricoh, 2026), and modest burn relative to Pinecone-class peers
Weaknesses
- Sub-scale revenue (~$12M ARR estimated for 2024) against well-funded rivals and free bundled alternatives
- No confirmed priced up-round since Apr 2023 while AI infra valuations re-rated - stale $200M mark cuts both ways
- Open-source core lets hyperscalers and users self-host without paying; conversion to paid cloud is the whole business
- Small team (~75-105 per trackers, possibly shrinking per the lowest 2026 estimate) spread across database engine, cloud platform, and agent tooling - execution breadth risk
Opportunities
- Agentic AI wave increases retrieval workloads per app — memory/context layers for agents are a natural extension
- Enterprise RAG platformization — compliance, hybrid deployment (BYOC/on-prem) where hyperscaler-native options are weak
- Consolidation in the vector DB category (Pinecone's reported 2025 sale exploration) could position Weaviate as a premium acquisition target or roll-up survivor
- Corporate/strategic channels (Ricoh's document-capture base) and vertical solutions (e-commerce search, financial research retrieval) monetizing above raw database consumption
Threats
- Commoditization — pgvector (free in Postgres), MongoDB Atlas Vector Search, Elasticsearch/OpenSearch vector, and hyperscaler-native stores erode the standalone vector DB category
- Pinecone, Qdrant, Milvus/Zilliz, Chroma, and Turbopuffer competing for the same greenfield AI workloads - Qdrant refreshed its war chest with a $50M Series B in Mar 2026
- Foundation-model context windows growing and retrieval moving into model/platform layers (OpenAI, Anthropic file/retrieval APIs) could shrink the independent retrieval market
- Down-round or distressed-exit risk if growth does not clear the bar set by 2023-era pricing - Elastic's CEO publicly argued in 2025 that standalone vector databases 'were never a business'
Moats, dependencies & bottlenecks
Moats
20M+ downloads, large integration ecosystem (LangChain, LlamaIndex); but OSS goodwill does not equal paid lock-in
Re-embedding and re-indexing production corpora plus rewritten query logic makes migration painful once embedded
multi-tenancy, agent tooling) Real today (Weaviate Agents shipped Mar 2025), but incumbents are closing the feature gap fast
Sub-scale vs hyperscalers and MongoDB; no meaningful data or network effects
Dependencies
Google Cloud GOOGL, Microsoft Azure MSFT) infrastructure + distribution Hosts Weaviate Cloud and runs the marketplaces it sells through - while all three ship competing vector offerings
Cohere, Anthropic, Hugging Face) ecosystem/technology Product value depends on tight model integrations; providers could absorb retrieval into their own APIs
Major top-of-funnel; framework defaults steer database choice
Pre-profitability; next priced raise likely required, priced against the 2023 mark; Ricoh strategic money (2026) extends runway but discloses no terms
Permissive license invites free-riding and cloud-provider forks; relicensing (Elastic/Redis path) would burn community trust
Advantages
- One of the few truly open-source (BSD-3), AI-native vector databases - deployable anywhere including air-gapped/regulated environments
- Native hybrid (vector + keyword) search and built-in vectorization modules reduce glue code vs bare vector stores
- Multi-tenancy architecture suited to SaaS builders running very large numbers of small isolated indexes (DocsBot case study cites 50,000+ tenants in one cluster)
- Capital-efficient operating posture extends runway and preserves optionality vs high-burn peers
Weaknesses
- Sub-scale revenue and no disclosed financials - opacity itself is a diligence risk
- Permissive OSS license means largest users may never pay
- Category pricing pressure from free (pgvector) and bundled (MongoDB, Elastic, hyperscaler) alternatives
- Stale 2023 valuation mark with no confirmed subsequent priced round
Bottlenecks
- OSS-to-paid-cloud conversion rate - the single constraint on revenue scale
- Enterprise sales capacity (~75-105 employees per trackers) vs land-grab pace of better-funded rivals
- Memory-heavy vector index economics (RAM cost) squeezing managed-cloud gross margin at large scale
- Differentiation bandwidth: must out-ship both pure-play rivals and platform incumbents simultaneously
Top signals & trends
Top signals
Either efficient burn or inability to command an up-round; contrast with rival Qdrant's $50M Series B (Mar 2026)
Corporate validation and a document/data-capture channel partner; but undisclosed terms reveal nothing about the valuation mark
Mirrors Series B terms exactly and is absent from Tracxn's funding-rounds page, Weaviate's investor page, and all tier-1 outlets - treated as a mislabeled duplicate of the Series B; if a flat round at 2023 pricing were real it would lean negative
Shipping cadence consistent with a healthy engineering org
Tracker data is noisy, but the range suggests disciplined-to-flat hiring, not a land-grab GTM build
Category-level signal that standalone vector DBs face exit-by-acquisition dynamics
Trends
More production LLM apps means more retrieval workloads; agent memory is an adjacent expansion
pgvector, MongoDB Atlas, Elastic, and hyperscaler-native stores absorb the low end of the market
Could shift retrieval value from independent databases to the model/platform layer
Favors open-source deploy-anywhere architectures over closed SaaS like Pinecone
Pinecone's 2025 sale exploration raises acquisition-exit probability for the category but can also compress standalone valuations
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Primary cloud infrastructure for Weaviate Cloud; AWS GenAI startup program participant
Cloud hosting and marketplace
Cloud hosting and marketplace
Embedding/LLM APIs consumed via built-in vectorizer modules
Embedding and reranking model integrations
Uses Weaviate in its Intelligence Engine Platform incl. the Mo research chatbot (weaviate.io case study)
Semantic search over 10M+ indexed posts (weaviate.io case study / Stack Overflow engineering posts)
AI document-processing platform processing 500K+ documents/day on Weaviate (weaviate.io case study)
DocsBot, Finster AI) Long tail of RAG/agent products on serverless cloud - the core revenue base
Closed-source managed vector DB; best-funded pure-play ($138M raised, $750M 2023 valuation); reportedly explored a sale via bankers in 2025 (The Information) after losing Notion as a customer
Bundles vector search into the document database enterprises already run - the strongest 'good enough' incumbent threat
Vector and hybrid search inside Elasticsearch; deep enterprise search installed base
Open-source (Rust) vector DB; ~$88M total raised after a $50M Series B (Mar 2026, led by Advance Venture Partners); closest like-for-like OSS rival, now better capitalized
Commercial steward of OSS Milvus, $113M raised; China-origin roots, now US-headquartered - competitive context only
Developer-favorite lightweight OSS vector store moving upmarket
Object-storage-based serverless vector search undercutting on cost; won marquee AI-native customers
Native vector options across its database portfolio plus Bedrock knowledge bases
Default retrieval layer for the Azure/OpenAI enterprise stack
GCP-native vector search, formerly Matching Engine
Free extension in the world's default OSS database; the strongest zero-cost substitute
Private; vector search bundled into the lakehouse where enterprise data already lives