Every agent session runs in its own sandbox. CPU and memory are metered by the second, tokens are passed through at the model provider's rates with no markup, and there is a 20% platform fee on the session's token and compute cost. There are no per-seat fees and nothing bills idle.

Is Ellipsis free for open source?

Open source projects are supported case by case; email team@ellipsis.dev to be set up with free usage.

Ellipsis Pricing in 2026: Plans, Limits and Real Costs

Ellipsis has no fixed tiers: it bills usage on Managed SaaS at tokens-at-cost plus $0.142/vCPU-hour, $0.024/GB-hour, and a 20% platform fee, with a one-time $100 credit to start. Solo founders should stay on Managed SaaS and skip the support packages.

Verified 2026-08-22 · official pricing page · Ellipsis in the directory

Quick pricing verdict

Starting price: $0 base; usage-based — tokens at cost, $0.142 / vCPU-hour, $0.024 / GB-hour, plus 20% platform fee. New organizations get a one-time $100 credit ($10 for individual accounts).
Free plan: No free tier. New organizations start with a one-time $100 credit ($10 for individual accounts) that is applied to usage before anything is billed.
Best plan for most people: Managed SaaS with no support package. Same features as the AWS deployment, no AWS setup, and the platform is fully self-serve.
Plan to avoid first: Premier support at $15,000 / month. It is an enterprise add-on, not required to use the platform.
Biggest hidden cost: The 20% platform fee added on top of token cost on every session, plus AWS charges if you choose the Your AWS Account deployment.
Pricing checked: 2026-08-22, against the official page

Ellipsis is priced as a single usage-based platform with two deployment paths: Managed SaaS (their cloud) and Your AWS Account (deployed into your VPC). Both run the full feature set with no per-seat fees and no idle charges; the bill is tokens at cost plus a 20% platform fee.

On Managed SaaS, compute is metered at $0.142 / vCPU-hour and $0.024 / GB-hour. On the AWS deployment, CPU, memory, and tokens are billed by AWS instead. Support is a separate, optional purchase — the page lists three enterprise packages from $5,000 to $15,000 per month.

Plans compared

Plan	Price	Best for	Key limits
Managed SaaS	$0 base + usage (tokens at cost, $0.142 / vCPU-hour, $0.024 / GB-hour, 20% platform fee)	Solo founders and teams that want to install the GitHub app and start without managing infrastructure.	No per-seat cap. $0 per-seat fees. Spending limits can be set at account, agent, session, and developer levels over rolling 1-, 7-, and 28-day windows; a session that would breach a limit never starts, and a running session that reaches its budget stops immediately.
Your AWS Account	$0 base + usage + AWS costs (20% platform fee; CPU, memory, and tokens billed by AWS)	Teams that need agents to run inside their own VPC with no credentials, code, or LLM calls leaving it.	Same agent configs and same prices as Managed SaaS. AWS metering replaces the $0.142 / vCPU-hour and $0.024 / GB-hour rates. $0 per-seat fees.
Standard support	$5,000 / month	Teams that want a direct Slack channel and guided onboarding.	First response time 48 hours. Shared support engineer. Optional add-on; not required to use the platform.
Advanced support	$7,500 / month	Teams that want faster response plus BYOC deployment help and security questionnaires / procurement support.	First response time 24 hours. Shared support engineer. Optional add-on; not required to use the platform.
Premier support	$15,000 / month	Enterprises that want a named engineer and priority feature requests.	First response time 4 hours. Named support engineer. Optional add-on; not required to use the platform.

Source: Ellipsis pricing page, checked 2026-08-22.

What each plan includes

Managed SaaS

Includes: Full feature set: code review, PR descriptions, screenshot attachments, dependency bump reviews, per-review and rolling daily/weekly review budgets, review analytics. Tokens at cost with no markup, $0.142 / vCPU-hour, $0.024 / GB-hour, 20% platform fee. $0 per-seat fees, $0 idle charges. One-time $100 credit for new organizations ($10 for individual accounts).
Missing: Nothing is gated on Managed SaaS — the page states both offerings run the full platform with no feature gates and no tiers.
Avoid if: You need data, code, and LLM calls to stay inside your own VPC; choose Your AWS Account instead.
Verdict: The default starting point for solo founders. Usage-based, fully self-serve, no support package required.

Your AWS Account

Includes: Same full feature set as Managed SaaS, deployable into your AWS VPC. No credentials, code, or LLM calls leave your VPC. Same per-agent configs, skills, and triggers. 20% platform fee. $0 per-seat fees.
Missing: Tokens, CPU, and memory are billed by AWS rather than at the Managed SaaS rates; specific AWS pricing is not shown on this page.
Avoid if: You don't already run on AWS or don't need VPC isolation; Managed SaaS is cheaper to start.
Verdict: Same platform, different billing rail. Plan the move with sales before migrating.

Standard support

Includes: Slack support channel, guided onboarding. First response time 48 hours. Shared support engineer.
Missing: Does not include named engineer, 24-hour response, BYOC deployment help, security questionnaires / procurement, or priority feature requests.
Avoid if: You only need basic usage — the platform is self-serve and a support package is not required to use it.
Verdict: Cheapest support tier; useful only if you want a direct Slack line and onboarding help.

Advanced support

Includes: Slack support channel, guided onboarding, help wiring BYOK (Anthropic or Bedrock), help with BYOC deployments, security questionnaires and procurement. First response time 24 hours. Shared support engineer.
Missing: Does not include named engineer, 4-hour response, or priority feature requests.
Avoid if: Your needs are covered by self-serve or Standard.
Verdict: Mid-tier support aimed at teams adopting BYOC and procurement workflows.

Premier support

Includes: Slack support channel, guided onboarding, BYOC deployment help, security questionnaires and procurement, priority feature requests. First response time 4 hours. Named support engineer.
Missing: Nothing on this page is excluded at Premier versus lower tiers.
Avoid if: You are a solo founder or small team; this is enterprise-tier pricing.
Verdict: Top-tier enterprise support; only worth it if a 4-hour SLA and named engineer matter to you.

Add-ons, overages, and surprise costs

Cost item	Trigger	Published amount	Billing behaviour
20% platform fee	Every agent session	20% of the session's token and compute cost	Added on top of token cost and (on Managed SaaS) on top of $0.142 / vCPU-hour and $0.024 / GB-hour.
AWS charges on Your AWS Account deployment	Choosing the Your AWS Account option	Billed by AWS for tokens, CPU, and memory instead of the Managed SaaS rates	Replaces the listed Managed SaaS CPU and memory rates; specific AWS amounts are not published on this page.
Support packages	Wanting a direct line to the Ellipsis team, onboarding help, or enterprise procurement support	$5,000 / month (Standard), $7,500 / month (Advanced), or $15,000 / month (Premier)	Optional and separate from usage billing; the page states a support package is not required to use the platform.

Source: Platform pricing docs, checked 2026-08-22.

Which plan should you choose?

Solo founder installing the GitHub app for the first time: Managed SaaS, no support package — Full feature set with no per-seat fees, $0 idle charges, and a one-time $100 credit to apply to first usage. Self-serve is explicitly supported.
Small team that needs code, credentials, and LLM calls to stay in its own VPC: Your AWS Account, no support package to start — Same agents, same configs, same prices as Managed SaaS, but deployed into your VPC. The page notes you can start on Managed SaaS and move later.
Procurement-led enterprise that wants a Slack channel, BYOC help, and security questionnaires: Advanced support ($7,500 / month) on top of either deployment — Adds BYOC deployment help, BYOK wiring for Anthropic or Bedrock, security questionnaires and procurement, and a 24-hour first response.

Frequently asked questions

How does the $100 credit work?: Every new organization starts with a one-time $100 credit ($10 for individual accounts). It applies to usage automatically before anything is billed.
How is usage billed?: Every agent session runs in its own sandbox. CPU and memory are metered by the second, tokens are passed through at the model provider's rates with no markup, and there is a 20% platform fee on the session's token and compute cost. There are no per-seat fees and nothing bills idle.
Do I need a support package to use Ellipsis?: No. The full platform is self-serve: install the GitHub app and your whole team is in. Support packages add a direct line to the team and enterprise deployment options on top.
Do I need to pay per developer?: No. Per-seat fees are $0 on both Managed SaaS and Your AWS Account. You only pay for what your agents do.
Can I set a spending limit?: Yes, at four levels — whole account, one agent, one session, and one developer — over rolling 1-, 7-, and 28-day windows. A session that would breach a limit never starts, and a running session that reaches its budget stops immediately.
Can I run Ellipsis in my own AWS account?: Yes. Agents are defined as YAML in your repositories, so the same configs, skills, and triggers run in either deployment at the same prices.
Is Ellipsis free for open source?: Open source projects are supported case by case; email team@ellipsis.dev to be set up with free usage.

What we read

Every figure on this page was extracted from the vendor's own pricing page at the moment below. The text is reproduced so you can check our reading against what they publish today.

Source captured 2026-08-22 08:13 UTC

Platform Use cases Pricing Enterprise Blog Documentation Sign in Install on GitHub Your bill = Tokens + 20% Pay for tokens (no markup), CPU/Memory (trivial), and a 20% platform fee. No per-seat fees, no idle charges. Same pricing in our cloud or your AWS account. Talk to sales Documentation Ask Claude about us Managed SaaS USAGE-BASED Every agent session runs in its own sandbox and you pay for exactly what it consumes. Your AWS Account USAGE-BASED + AWS COSTS Deploys into your AWS VPC. No credentials, code, or LLM calls leave it. Per-seat fees $0 $0 Tokens At cost, no markup Billed by AWS Platform fee 20% of token cost 20% of token cost CPU $0.142 / vCPU-hour Billed by AWS Memory $0.024 / GB-hour Billed by AWS Control plane Ellipsis cloud Your AWS VPC Agent sandboxes Ellipsis cloud Your AWS VPC Feature set Full Full Start on Managed SaaS and move to your own AWS account later. Same agents, same configs. Everything included Both offerings run the full platform. No feature gates, no tiers. Agents Code review Platform Models One toggle to turn on. Custom review agents defined in code_review.yaml. Incremental by design. Reviews only new commits, never re-comments. PR descriptions. Kept current automatically on every push. Screenshot attachments. Agents persist evidence onto PRs. Reviews bot PRs. Dependency bumps reviewed out of the box. Review budgets. Per-review caps plus trailing daily and weekly limits. Review analytics. Signal rate, severity mix, and cost per posted comment. Support Packages You don't need a support package to use Ellipsis. The platform is fully self-serve; these are for enterprises that want a direct line to our team. Features Standard $5,000 / month Advanced $7,500 / month Premier $15,000 / month Slack support channel Guided onboarding Help building custom agents Self guided BYOC Help wiring BYOK (Anthropic or Bedrock) Help with BYOC deployments — Security questionnaires and procurement — Priority feature requests — — First response time 48 hours 24 hours 4 hours Support engineer Shared Shared Named See Ellipsis in action Walk through the platform with our team and get your first agents deployed. Talk to sales Frequently asked questions How does the $100 credit work? Every new organization starts with a one-time $100 credit ($10 for individual accounts). It applies to your usage automatically before anything is billed. How much does it cost? Every agent session runs in its own sandbox, and you pay for exactly what that session consumes: CPU and memory metered by the second, tokens passed through at the model provider's rates with no markup, and a platform fee of 20% of the session's token and compute cost. There are no per-seat fees, and nothing runs idle, so nothing bills idle. Do I need a support package to use Ellipsis? No. The full platform is self-serve: install the GitHub app and your whole team is in. Support packages add a direct line to our team and enterprise deployment options on top. Do I need to pay per developer? No. There are no per-seat fees. Your entire team can use Ellipsis. You only pay for what your agents do. Is it free for open source? We support open source projects case by case. If you maintain one, email team@ellipsis.dev and we can set you up with free usage. Can I set a spending limit? Yes, at four levels: your whole account, one agent, one session, and one developer, over rolling 1-, 7-, and 28-day windows. A session that would breach a limit never starts, and a running session that reaches its budget stops immediately. Can I run Ellipsis in my own AWS account? Yes. Agents are defined as YAML in your repositories, so they are portable by construction: the same configs, skills, and triggers run in either deployment, at the same prices. Talk to sales and we will help you plan the move. Is there enterprise pricing? Yes. The support packages above cover Slack support, Bring Your Own Cloud deployments, guided onboarding, and priority response times. Email team@ellipsis.dev to discuss.

Pricing change log

What moved since we started tracking this page, newest first. A price that changes silently is the failure mode of every other pricing guide.

2026-08-22First published, extracted from https://www.ellipsis.dev/pricing.