IT AI Index
September 2026 Edition · The permanent record of this edition. The unqualified address always carries the latest edition.
Index › Data platform › September 2026 Edition

Vector databases

Asked as “vector database”, and as “vector search database for AI applications”, on behalf of a mid-market B2B company. 51 first choices recorded across the direct, paraphrase, budget and scale prompts, twelve models each.
Standing
Qdrant
Contested
37% of first choices, contested.
ShowinginAll 67 categories in the September 2026 Edition →

By buyer segment

The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.
Small businesspgvector46%Clear leaderStanding →Mid-marketQdrant37%ContestedThis pageEnterprisePinecone53%Clear leaderStanding →

01The standing

Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all twelve models, for a mid-market B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
01Qdrant37%5%63endorsed leader
02pgvector35%14%51endorsed leader
03Pinecone14%28%58criticized challenger
04Weaviate6%15%61accepted challenger
05Chroma4%27%30criticized challenger
06Milvus2%37%38criticized challenger
Show the two products at 0%, ordered by negative rate
08Milvus / Zilliz Cloud0%36%14criticized challenger
07FAISS0%18%11accepted challenger
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its vendor page.
Head to head
Qdrant vs pgvectorQdrant vs Pineconepgvector vs Pinecone

Recommended versus criticized

Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.

Criticized challengerCriticized default
Negative label rate →
01
02
03
04
05
06
F07
08
Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative50%
Key
01Qdrant37%
02pgvector35%
03Pinecone14%
04Weaviate6%
05Chroma4%
06Milvus2%
07FAISS0%
08Milvus / Zilliz Cloud0%

02What they warned about

One of twelve models held their first choice under the paraphrase. Claude Haiku 4.5, GPT-5.4 mini, Perplexity Sonar, Grok 4.1 Fast, Mistral Small, DeepSeek V4 Flash, Llama 4 Maverick, Qwen 3.7 Flash, Kimi K2, GLM 4.7 FlashX and MiniMax M2.5 changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Pinecone
28%
16 of 58 labels negative · 11 of 12 models · 5 hard negative
“Specific Targets to Watch Out For:** **Pinecone** (purely managed serverless)... you cannot take your data elsewhere without significant re-architecture.” Qwen 3.7 Flash, negative prompt
Milvus
37%
14 of 38 labels negative · 8 of 12 models · 1 hard negative
“What to Avoid (For Now)... Milvus / Vespa... generally overkill for mid-market requirements.” Gemini 3.5 Flash, direct prompt
Weaviate
15%
9 of 61 labels negative · 8 of 12 models · 1 hard negative
“**Avoid Initially:** Pinecone/Weaviate Cloud (managed convenience costs $25-70/mo min after free tiers” Grok 4.1 Fast, budget prompt
pgvector
14%
7 of 51 labels negative · 7 of 12 models
“When you need very high-scale, pure vector search | It is often recommended for existing Postgres shops, but can be a weaker fit for large-scale semantic search workloads.” Perplexity Sonar, negative prompt

03What they cite

Citations exist only for the models that return a source list: twelve of the twelve in this edition, and all six flagship models on the expanded tier.

Sites the answers cite

68 of 72 answers in this category came back with a source list, from 12 of 12 models: citations where the model returns them, or the search results it consulted. 6 of those lists are Google grounding redirects that name no site and are left out of the counts. 804 links across 243 sites, every framing counted. Ranked by the number of answers carrying the site or page.

firecrawl.dev
42 answers · 42 citations · 10 models
encore.dev
26 answers · 26 citations · 9 models
dev.to
24 answers · 30 citations · 9 models
redis.io
vendor site · Redis18 answers · 19 citations · 9 models
zenml.io
vendor site · ZenML16 answers · 16 citations · 7 models
marktechpost.com
14 answers · 15 citations · 7 models
g2.com
14 answers · 14 citations · 7 models
iternal.ai
14 answers · 14 citations · 7 models
actian.com
12 answers · 16 citations · 10 models
medium.com
12 answers · 16 citations · 7 models
liveblocks.io
12 answers · 12 citations · 7 models
pecollective.com
12 answers · 12 citations · 7 models

Pages the answers cite

The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.

firecrawl.dev/blog/best-vector-databases
42 answers · 10 models
encore.dev/articles/best-vector-databases
26 answers · 9 models
zenml.io/blog/vector-databases-for-rag
16 answers · 7 models
g2.com/articles/best-vector-databases
14 answers · 7 models
iternal.ai/insights/best-vector-databases-2026
14 answers · 7 models
liveblocks.io/blog/whats-the-best-vector-database-for-building-ai-products
12 answers · 7 models
redis.io/blog/best-open-source-vector-databases-comparison
12 answers · 9 models
actian.com/blog/databases/how-to-evaluate-vector-databases-in-2026
11 answers · 9 models
atlan.com/know/top-vector-databases-enterprise-ai
11 answers · 8 models
marktechpost.com/2026/05/10/best-vector-databases-in-2026-pricing-scale-lim…
10 answers · 5 models

04How they answered

Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ModelDirect“What is the best vector database for a mid-market B2B company?”Paraphrase“Which vector search database for AI applications would you recommend to a mid-sized B2B company?”Comparative“What are the top vector databases and how do they differ?”Budget-constrained“What is the best vector database for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating a vector database. What should we look at?”Negative“Which vector databases should I avoid or be cautious about?”
Claude Haiku 4.5Pinecone, Qdrant
Two alternativesWeaviate, pgvector
against: MilvusPineconeChanged
Two alternativesChroma, pgvector
no first choiceChroma
Three alternativesMilvus, Qdrant, Weaviate
no first choiceagainst: Chroma, Milvus, Pinecone, Qdrant, Weaviate
GPT-5.4 miniQdrant, pgvector
Three alternativesMilvus / Zilliz Cloud, Pinecone, Weaviate
QdrantChanged
Three alternativesMilvus, Pinecone, Weaviate
no first choice
Five alternativesElasticsearch / OpenSearch, Milvus, Pinecone, Weaviate, pgvector
Qdrant
Two alternativesChroma, Milvus
against: Pinecone, Weaviateno first choicenothing named
Gemini 3.5 Flashpgvector
Three alternativesPinecone, Qdrant, Weaviate
against: Chroma, Milvus, VespapgvectorHeld
Three alternativesPinecone, Qdrant, Weaviate
pgvector
Six alternativesChroma, LanceDB, Milvus, Pinecone, Qdrant, Weaviate
pgvector
Five alternativesChroma, LanceDB, Pinecone, Qdrant, Turbopuffer
against: Milvus, Weaviatepgvector
Three alternativesPinecone, Qdrant, Weaviate
against: Milvusagainst: Chroma, FAISS, Milvus, Pinecone, pgvector
Perplexity SonarQdrant, pgvector
Three alternativesMilvus, Pinecone, Weaviate
QdrantChanged
Three alternativesPinecone, Weaviate, pgvector
no first choiceQdrant, pgvector
One alternativeChroma
against: Pineconeno first choiceagainst: Elasticsearch, Milvus, MongoDB Atlas Vector Search, Pinecone, Qdrant, Weaviate, pgvector
Grok 4.1 Fastpgvector
Two alternativesPinecone, Qdrant
against: Chroma, Milvus / Zilliz CloudQdrant, pgvectorChanged
Two alternativesPinecone, Weaviate
against: Annoy, FAISS, Milvus / Zilliz Cloudno first choiceChroma
Two alternativesQdrant, pgvector
against: Pinecone, WeaviatePinecone
Four alternativesMilvus / Zilliz Cloud, Qdrant, Weaviate, pgvector
against: Chromaagainst: ChromaDB, Milvus, Pinecone, Weaviate
Mistral Smallpgvector
Three alternativesPinecone, Qdrant, Weaviate
MilvusChanged
Three alternativesElasticsearch, MongoDB Atlas Vector Search, Weaviate
no first choiceQdrant
Three alternativesChromaDB, Weaviate, pgvector
no first choiceagainst: ChromaDB, Milvus / Zilliz Cloud
DeepSeek V4 FlashQdrant, Weaviate
One alternativepgvector
against: Milvus, PineconeQdrantChanged
Three alternativesPinecone, Weaviate, pgvector
against: Milvus / Zilliz Cloudno first choiceQdrant, pgvector
One alternativeChromaDB
against: Milvus / Zilliz Cloudno first choice
One alternativepgvector
against: ChromaDB, Elasticsearch, Milvus, MongoDB Atlas Vector Search, Pinecone, Redis, Weaviate
Llama 4 Maverickpgvectorno first choiceChangedno first choice
Five alternativesChroma, Milvus, Pinecone, Qdrant, Weaviate
Qdrantagainst: Milvus, Pineconeno first choiceagainst: Pinecone, Qdrant, Weaviate, pgvector
Qwen 3.7 Flashpgvector
Three alternativesPinecone, Qdrant, Weaviate
against: MilvusPinecone, WeaviateChanged
Two alternativesQdrant, pgvector
Pinecone
Four alternativesMilvus / Zilliz Cloud, Qdrant, Weaviate, pgvector
against: ChromaDBQdrant
Three alternativesChroma, Milvus, pgvector
no first choiceagainst: Milvus, Pinecone, Zilliz Cloud, pgvector
Kimi K2Pinecone, pgvector
Two alternativesQdrant, Weaviate
against: ChromaPineconeChanged
Four alternativesChroma, Qdrant, Weaviate, pgvector
Pinecone
Five alternativesChromaDB, Milvus / Zilliz Cloud, Qdrant, Weaviate, pgvector
Qdrant, pgvector
Two alternativesChroma, Milvus
Pinecone
Five alternativesMilvus, Qdrant, Weaviate, Zilliz Cloud, pgvector
against: Chroma, Pinecone, pgvector
GLM 4.7 FlashXpgvector
Three alternativesPinecone, Qdrant, Weaviate
QdrantChanged
Four alternativesMilvus / Zilliz Cloud, Pinecone, Weaviate, pgvector + pgvectorscale
Pinecone
Seven alternativesChromaDB, LanceDB, Milvus / Zilliz Cloud, Qdrant, Vespa, Weaviate, pgvector + pgvectorscale
against: Elasticsearch / OpenSearchQdrant, pgvector
Four alternativesChromaDB, Milvus, Redis, Weaviate
against: Pineconeno first choiceagainst: Pinecone, Redis, Weaviate, pgvector
MiniMax M2.5pgvector
Three alternativesPinecone, Qdrant, Weaviate
QdrantChanged
Two alternativesPinecone, pgvector
no first choice
Seven alternativesChroma, Milvus / Zilliz Cloud, Pinecone, Qdrant, Vespa, Weaviate, pgvector
Qdrant, Weaviate
One alternativeChroma
against: Milvuspgvector + pgvectorscale
Five alternativesMilvus, Pinecone, Qdrant, Weaviate, Zilliz Cloud
against: Chroma, Pinecone, pgvector
Bold is the first choiceAlternatives are counted; the count opens them.What the answer argued against

05The record

One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Seventy-two rows: every prompt, every model, every answer.
PromptModelVersion stringTime (UTC)SearchedSourcesLatency
Direct recommendationClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 10:12yes179 s
Direct recommendationGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 12:16yes16 s
Direct recommendationGemini 3.5 Flashgemini-3.5-flash2026-09-17 10:27yes925 s
Direct recommendationPerplexity Sonarsonar2026-09-17 10:31yes206 s
Direct recommendationGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 10:14yes2917 s
Direct recommendationMistral Smallmistral/mistral-small via mistral2026-09-17 11:30yes54 s
Direct recommendationDeepSeek V4 Flashdeepseek/deepseek-v4-flash via fireworks2026-09-17 09:51yes1718 s
Direct recommendationLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 09:57yes52 s
Direct recommendationQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 09:38yes529 s
Direct recommendationKimi K2moonshotai/kimi-k2 via novita2026-09-17 12:12yes1022 s
Direct recommendationGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 09:33yes2856 s
Direct recommendationMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 10:46yes918 s
ParaphraseClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 10:41yes97 s
ParaphraseGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 13:42yes45 s
ParaphraseGemini 3.5 Flashgemini-3.5-flash2026-09-17 10:12yes1123 s
ParaphrasePerplexity Sonarsonar2026-09-17 14:07yes205 s
ParaphraseGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 10:06yes3316 s
ParaphraseMistral Smallmistral/mistral-small via mistral2026-09-17 13:52yes56 s
ParaphraseDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-09-17 13:26yes2131 s
ParaphraseLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 13:51yes53 s
ParaphraseQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 12:58yes533 s
ParaphraseKimi K2moonshotai/kimi-k2 via novita2026-09-17 11:21yes1448 s
ParaphraseGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 11:04yes1424 s
ParaphraseMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 11:53yes1040 s
ComparativeClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 11:12yes1811 s
ComparativeGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 12:02yes68 s
ComparativeGemini 3.5 Flashgemini-3.5-flash2026-09-17 14:08yes1825 s
ComparativePerplexity Sonarsonar2026-09-17 10:30yes209 s
ComparativeGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 10:19yes2814 s
ComparativeMistral Smallmistral/mistral-small via mistral2026-09-17 13:23yes913 s
ComparativeDeepSeek V4 Flashdeepseek/deepseek-v4-flash via fireworks2026-09-17 11:16yes1820 s
ComparativeLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 12:10yes53 s
ComparativeQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 12:19yes526 s
ComparativeKimi K2moonshotai/kimi-k2 via novita2026-09-17 13:01yes837 s
ComparativeGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 11:20yes18222 s
ComparativeMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 12:06yes517 s
Budget constrainedClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 09:26yes98 s
Budget constrainedGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 13:14yes45 s
Budget constrainedGemini 3.5 Flashgemini-3.5-flash2026-09-17 10:51yes1418 s
Budget constrainedPerplexity Sonarsonar2026-09-17 12:07yes206 s
Budget constrainedGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 10:31yes1616 s
Budget constrainedMistral Smallmistral/mistral-small via mistral2026-09-17 12:33yes86 s
Budget constrainedDeepSeek V4 Flashdeepseek/deepseek-v4-flash via fireworks2026-09-17 10:08yes912 s
Budget constrainedLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 12:27yes73 s
Budget constrainedQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 10:07yes526 s
Budget constrainedKimi K2moonshotai/kimi-k2 via novita2026-09-17 10:50yes517 s
Budget constrainedGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 11:45yes1450 s
Budget constrainedMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 13:39yes525 s
Scale constrainedClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 10:43no07 s
Scale constrainedGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 12:12no08 s
Scale constrainedGemini 3.5 Flashgemini-3.5-flash2026-09-17 10:16yes927 s
Scale constrainedPerplexity Sonarsonar2026-09-17 10:27yes2010 s
Scale constrainedGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 11:40yes1510 s
Scale constrainedMistral Smallmistral/mistral-small via mistral2026-09-17 12:13no07 s
Scale constrainedDeepSeek V4 Flashdeepseek/deepseek-v4-flash via fireworks2026-09-17 11:59yes1027 s
Scale constrainedLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 11:19yes52 s
Scale constrainedQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 10:31no042 s
Scale constrainedKimi K2moonshotai/kimi-k2 via novita2026-09-17 11:46yes1326 s
Scale constrainedGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 10:14yes2065 s
Scale constrainedMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 14:05yes525 s
Negative framingClaude Haiku 4.5claude-haiku-4-5-202510012026-09-17 10:33yes1510 s
Negative framingGPT-5.4 minigpt-5.4-mini-2026-03-172026-09-17 09:57yes68 s
Negative framingGemini 3.5 Flashgemini-3.5-flash2026-09-17 10:51yes1528 s
Negative framingPerplexity Sonarsonar2026-09-17 09:25yes2010 s
Negative framingGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-09-17 10:48yes4120 s
Negative framingMistral Smallmistral/mistral-small via mistral2026-09-17 10:25yes54 s
Negative framingDeepSeek V4 Flashdeepseek/deepseek-v4-flash via fireworks2026-09-17 14:06yes2824 s
Negative framingLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-09-17 11:30yes53 s
Negative framingQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-09-17 10:24yes950 s
Negative framingKimi K2moonshotai/kimi-k2 via novita2026-09-17 12:38yes3036 s
Negative framingGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-09-17 11:00yes1465 s
Negative framingMiniMax M2.5minimax/minimax-m2.5 via minimax2026-09-17 10:11yes1529 s

Normalization in this category

Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.

Category-scoped readings
MongoDB read as MongoDB Atlas Vector Search
MongoDB Atlas read as MongoDB Atlas Vector Search
PostgreSQL read as pgvector
Unresolved, counted raw
Actian VectorAI
Elasticsearch with ANN
Discontinued, still offered
No shut-down product was recommended here.

Also in data platform

The other categories of this function, with the leader in each.
← StreamingOrchestration →