Key findings
- Vendr's data on 174 purchases (as of Oct 28, 2026) puts the median Glean contract at US$98,890 a year, ranging from US$29,880 to US$208,897, and minimums often start at 100 to 250 users.
- Glean documents two pricing models. Enterprise Flex is per-user seats plus a pooled FlexCredit allowance. Glean Core Suite is per-user seats plus Model Hub usage at provider API rates against an upfront commit.
- An Enterprise Flex seat includes everyday search and fast-mode chat on Basic and Standard models. Premium models, agent runs, slides, and meeting notes consume credits, and learned memories start consuming credits on November 7, 2026.
- Under Model Hub usage, Glean's own Waldo model lists at US$0.50 input and US$2.50 output per million tokens, and other models bill at their provider's API rates.
- Of 51 public accounts, 6 discuss price and 3 raise it as a concern.
Budget Glean as seats plus pooled credits; premium models and agents add usage spend on top of seats. Glean publishes no seat price; competitor Slite reported the one public per-seat figure in an April 2026 deployment. For a few hundred users, Slite reported about US$45 to US$48 per user per month.
What Glean publishes
Glean's pricing URL redirects to its home page, and its product pages send buyers to Glean's sales team. Its documentation does describe both current pricing models in detail, updated October 2 and October 5, 2026.
- Enterprise Flex: per-user, per-month seats designed for every employee, with a base pool of FlexCredits shared across the organization and FlexCredit Packs to add more.
- Glean Core Suite: per-user, per-month seats covering search, the Assistant, agent building, connectors, Protect, and hosting, with Model Hub usage drawn from an upfront commit at model or hosting provider API rates.
- Flexible Model Management under Core Suite: a fee set as a share of total LLM usage or of the upfront commit.
- Site-wide annual fees: Glean Protect+ and Premium Support.
What a seat includes, and what draws on credits
On Enterprise Flex, basic search, fast-mode Assistant queries, Collections, Go Links, the people directory, and agent creation are unlimited. Thinking and Adaptive Reasoning queries on Standard models are included up to 100 per user per week.
Glean's rate card gives typical consumption at the 50th and 90th percentile. An agent run uses about 7 and 114 FlexCredits. A Premium-model Thinking query uses about 35 and 120. Slide generation uses about 45 and 142. Learned memories begin billing November 7, 2026.
Discounts and model choice
Seats and FlexCredits are discounted for customers who supply their own LLM keys or host Glean in their own GCP or AWS account, to offset the compute and model costs they carry. Customer Hosted deployments can count against existing cloud spend commitments.
Under Model Hub usage, Glean bills successful LLM calls only, at listed rates. Per million tokens, Glean Waldo costs US$0.50 input and US$2.50 output. Provider models include Claude Opus 5.5 at US$4 and US$20 and GPT 6 Sol at US$2 and US$10.
What buyers report paying
Vendr, which negotiates software purchases for its customers, reports that Glean buyers save 20% on average (as of Oct 28, 2026) and that two- to three-year terms are common. Slite, which sells a competing knowledge tool, reports one April 2026 deployment at about 200 to 230 users, and a company of about 700 employees that estimated US$250,000 a year. Metronome's pricing index lists Premium Support at 12% of license fees, which it marks as confirmed.
After an intro meeting, a Hacker News commenter judged the pricing to make sense only at 100 to 150 employees or more and recalled a US$50k yearly starting point. The commenter cautioned that pricing may have changed. A senior support engineer found licenses expensive and hard to justify internally. Another commenter found unified search helpful but expensive.
A principal architect at a 10,001+ employee tech vendor calls the price fair for the features. A Databricks recruiter and a Carta account executive, reviewing on TrustRadius, both answered that they were not sure Glean is good value for the price.
Methodology
GILD read Glean's Core Suite and Enterprise Flex pricing documentation, usage controls page, and deployment documentation on October 12, 2026. Counts of user opinion come from 51 public accounts published from January 1, 2025 to September 22, 2026. On October 25, 2026 GILD added contract figures from Vendr's procurement data, a buyer report from Slite, and Metronome's pricing index. Slite sells a competing product, and Metronome marks its seat figures as third-party estimates, so GILD cites only its confirmed support fee.
Products in this piece
Sources
- Official documentation
- 6
- Public source
- 8
- Glean docs: Glean Enterprise Flex pricing
- Glean docs: Glean Core Suite pricing
- Glean: AI usage controls
- Glean docs: Usage dashboard for Enterprise Flex
- Glean docs: Glean deployment models
- Glean: Enterprise AI that works
- Hacker News comment on Glean pricing after an intro meeting (Jan 2, 2025)
- PeerSpot review by a senior support engineer: Unified data search (Mar 16, 2026)
- Hacker News comment on unifying search tools and their cost (Aug 27, 2026)
- PeerSpot review by Gopi Gonuguntla, principal architect (Mar 14, 2026)
- TrustRadius review by Savannah Hair, Carta: Glean vs. Guru (Nov 5, 2025)
- Vendr: Glean Software Pricing & Plans 2026 (updated Feb 2026)
- Slite: Glean Review, Should it be your Enterprise Search tool? (Ishaan Gupta, Jun 2, 2026)
- Metronome: Glean pricing index (viewed Oct 25, 2026)