Subcategory · AI Citation Index
LLM Hosting
Hugging Face leads LLM hosting at 59% shortlist rate; specialized platforms Replicate and Together AI trail in the low 40s.
195 discovery queries · 289 head-to-heads · refreshed Oct 8, 2026
Discovery stage
The shortlist
Across 195 buyer-style "LLM Hosting" queries
Hugging Face appears in 116 of 195 prompts (59.5% shortlist rate) and generates 246 mentions across four models, making it the category anchor. Replicate (43.6%) and Together AI (42.6%) follow, while hyperscalers Google Cloud and Databricks hover near 40% but lack the pure-play focus. Fireworks AI and Anyscale sit in the mid-20s to low-30s, and nine brands total compete for mindshare.
Hover or click a logo to see brand details
Get weekly AI visibility changes for LLM Hosting sent to your inbox.
Score shifts, new entrants, citation gaps — every Monday.
Signal by intent
By topic
Top 5 most-cited brands per intent cluster. Brands with zero citations in a topic are not shown.
Evaluation stage
Head-to-head
How often AI cites each brand across uniform category evaluation prompts · median 6/100
Together AI scores highest in head-to-head comparisons (43 avg across 34 queries), edging Hugging Face (41, 33 queries) and Replicate (40, 32 queries). Databricks (38) and Fireworks AI (35) follow, while Azure OpenAI Service, Google Cloud, and Anyscale all land at 28–29. The median category score of 6 suggests most brands face narrow evaluation windows or niche use cases.
Hover or click a logo to see brand details
Each brand's score is the share of category evaluation prompts where AI cited them across all four engines — the same prompt pool for every brand. Brands above the median citation rate have stronger presence in evaluation-stage queries.
Brands to know
In this category
Hugging Face
Dominates shortlists with a 60% appearance rate and 246 mentions. Its model-hub roots give it unmatched developer visibility in LLM hosting discussions.
Read brand profile →Anyscale
Scores 56 despite a 29.7% shortlist rate, indicating strong performance in Ray-native workloads. Appears in 58 prompts but earns only 78 mentions, signaling narrow but loyal consideration.
Read brand profile →Replicate
43.6% shortlist rate and 152 mentions position it as the go-to for API-first model serving. Balances breadth (85 prompts) with consistent evaluation scores (40 avg).
Read brand profile →Together AI
Leads evaluation queries (43 avg, 34 comparisons) but trails in shortlist share (42.6%). High mention count (170) suggests frequent context but less decisive wins.
Read brand profile →Modal
Scores 52 from a 25.6% shortlist base, pointing to specialized infrastructure appeal. 61 mentions across 50 prompts indicate targeted developer workflows rather than broad adoption.
Read brand profile →Citation sources
Where AI pulls citations from
804 citations captured across LLM Hosting prompt runs.
Vendor pages
189Product, help, and marketing pages from tracked vendors
Independent sources
311Reviews, encyclopedias, forums, press — not vendor-owned
Buyer questions
What AI cites for top LLM Hosting questions
Most-cited prompts across the buyer journey. Click any prompt to see the actual URLs AI engines link to.
Discovery
Buyers exploring the category- LLMOps Course | Deploy & Scale Production LLMsschoolofcoreai.com
- The Roadmap for Mastering Language Models in 2025 - MachineLearningMastery.commachinelearningmastery.com
Evaluation
Buyers comparing options- Self Hosting LLMs in June 2025 - Matt Ferranteheyferrante.com
- Best enterprise LLMs for accuracy and business use in 2025techresearchonline.com
Want to know if AI cites your brand for LLM Hosting?
Free audit. ChatGPT, Perplexity, Gemini, Claude.
Run an audit →