--- title: "Pricing · SourceAnt" description: "Free for 250 events a month, then plans from $49, priced on repositories and analysis rather than seats. Bring your own model keys and we bill you nothing for inference." url: "https://sourceant.ai/pricing/" --- Free for 3 repositories and 250 events a month, no card # Start free. One price for the whole team. Every workspace starts on Free. When you need more, pick a band of repositories and monthly analysis, then connect your own provider key and we bill you nothing for inference. Add every engineer and every agent you have; the platform price does not move. Run it inside your own network if your code cannot leave it. Bring your own keys and we bill you nothing for inference: your provider invoices you directly. The platform price is the same either way. ## Free $0 per month, no card 3 repositories · 250 events / month then it pauses until the first of the month Models: nothing from us Models: cost plus 10%, itemised The whole platform for a small team or a side project: about 10 pull requests a month, never billed past the allowance. - Local code, knowledge, and system graphs over MCP - Durable context and memory, with proposal curation - Contract analysis and continuous indexing at scale - Requirement records, links, and coverage - The dashboard, workspaces, and team roles - Lens, the review reading surface - Hosted MCP serving and analytics - Unlimited seats and agents [Start free](https://app.sourceant.ai/) ## Starter $49 per month, platform Up to 25 repositories · 5,000 events / month then $8 per extra 1,000 events Models: nothing from us Models: cost plus 10%, itemised For a team that needs more than the free allowance and cannot have reviews pause. - Everything in Free - Never pauses: extra events bill per 1,000 - Unlimited seats and agents [Start free](https://app.sourceant.ai/) Most popular ## Growth $249 per month, platform Up to 150 repositories · 40,000 events / month then $6 per extra 1,000 events Models: nothing from us Models: cost plus 10%, itemised For an engineering org shipping heavily with agents across a real estate of services. - Everything in Starter - Priority queueing when reviews queue up - Unlimited seats and agents - Email support [Start free](https://app.sourceant.ai/) ## Scale $1,200 per month, from, platform Unlimited repositories · 200,000 events / month then $3 per extra 1,000 events Models: nothing from us Models: cost plus 10%, itemised For a large org where the graph is the thing everyone builds against. Priced by band, never by headcount. - Everything in Growth - Dedicated graph capacity and higher query limits - SSO and audit log - Unlimited seats and agents - Priority support and a named contact [Talk to us](mailto:hello@sourceant.ai?subject=SourceAnt%20Scale) ## On-premise For teams whose code never leaves the building. The full platform deployed inside your network, air-gapped if you need it, with your own model endpoints, premium plugins included and no event metering. [Talk to us](mailto:hello@sourceant.ai?subject=SourceAnt%20on-prem&body=We%20are%20interested%20in%20running%20SourceAnt%20on-prem.%0A%0ATeam%20size%3A%0AInfrastructure%3A%0AAir-gapped%3A%20yes%2Fno) ## Work out your month Put in what your team ships in a typical month. Nothing is sent anywhere; the sum happens in your browser. Pull requests reviewed a month Issues triaged a month Commits pushed a month 1,000 events a month Lands on Starter $49 well inside the 5,000 included ### What each one weighs - A pull request reviewed one verdict against the graph, however many files it touches - An issue triaged routed and enriched from the code it names - A push indexed per commit that changes tracked code - A context or contract captured whether proposed by us or written by you - An MCP query answered 0 never metered: reading a graph you already own should not be rationed A monorepo counts as one repository, and a quiet one costs you nothing but its place in the band. Splitting a service out should be an architecture decision, not a billing one. ## And what the models cost The plan price is the platform: holding your repositories, indexing them, and serving the graph. That costs us the same whoever pays for inference, so it is the same price either way. Models are separate, and the choice is yours on every plan. Bring your own keys no charge from us connect an OpenAI, Anthropic, or compatible key; you pay your provider directly and we never see the bill Cloud models cost plus 10% every token consumed, itemised, on any plan, with no commitment On-premise always uses your own keys, because the models are called from inside your network. ## Two ways to start Run the engine on your own machine, or connect a repository on the cloud and read the first review. Neither needs a card. On your machine `curl -fsSL https://raw.githubusercontent.com/sourceant/cli/main/scripts/install.sh | sh` [Start on the cloud](https://app.sourceant.ai/) [Read the install guide](https://sourceant.ai/docs/local-setup) ## Questions ### How the price works ### Do you charge per seat? No, and we do not intend to. Adding engineers or agents does not change your bill. Plans scale on the two things that actually cost us money: how many repositories we hold in the graph, and how much work we do against them each month. ### What counts as an event? A unit of analysis we run for you. A pull request reviewed is 10, an issue triaged is 5, a push indexed or a context captured is 1. Reading the graph over MCP is free and always will be, because querying what you already own should not be rationed. ### Why bands rather than pure usage? A pure meter makes the bill unpredictable and punishes exactly the teams doing the most with agents. A band gives you a number you can budget, with overage only if you go past it. Most teams never do. ### Is the overage rate for model usage? No. Overage is the platform meter: events past your included allowance, charged whether or not you bring your own keys. Model usage is separate and is only on your bill if we are running the models for you. ### Why does overage get cheaper on bigger plans when model usage does not? Because our costs behave differently. Platform cost per event falls with volume: fixed infrastructure amortises and storage is bought in bulk, so a bigger commitment earns a lower marginal rate. Inference does not work that way. We pay roughly provider list whatever the volume, so cloud models stay at cost plus 10 percent on every plan. Tiering that would mean inventing a discount we do not actually get. ### Is there a free plan? Yes. Every workspace starts on Free: 250 events a month across 3 repositories, with every feature and no card. That is 25 reviews, about 10 pull requests, because a pull request is reviewed again when new commits are pushed to it. It stops at the allowance rather than billing past it, and starts again on the first of the month. ### What happens if we go over a band? On a paid plan, nothing stops. Events past the included allowance bill at the overage rate for your plan, and we tell you when you cross so it is never a surprise at the end of the month. ### Is it ever cheaper to stay on a smaller plan and pay overage? No. Every overage rate is set above what the next plan charges for the same volume, so by the time you would reach the next band, moving up already costs less. We would rather you paid us less than felt trapped on a plan you had outgrown, so we tell you when you cross the line. ### We have thousands of engineers on a handful of monorepos. What do we pay? Whatever band your event volume lands in, and nothing for the engineers. A monorepo counts as one repository, so a large team on few repositories usually sits lower than its headcount would suggest. Talk to us and we will size it from your real pull-request and commit volume rather than guessing. ### Models and keys ### Does bringing our own model keys reduce the plan price? No, and that is deliberate. The plan price is the platform: repositories held, indexed, and served. That costs us the same whoever pays for inference, so charging you less for it would just mean charging someone else more. What bringing your own keys removes is the model bill on top, which is the part that actually scales with how hard you use us. ### Can we bring our own keys on any plan? Yes, including the cheapest one. Our buyers tend to have provider contracts, committed spend, or rules about which models may see their code, and none of that should force an upgrade. Connect a key and we bill you nothing for inference. ### Are retried runs billed? Yes. You are billed for the tokens actually consumed, and a run that had to be retried consumed them twice. It is itemised, so a retry is visible on the invoice rather than folded into a rounded-up total. Bringing your own keys does not change this either: your provider bills the same tokens, they just bill you directly. ### What do cloud models cost? What the provider charged us plus 10 percent, itemised on your invoice so you can check it against their pricing page. The margin is deliberately thin: you are already paying for the platform, and we would rather run the models as a convenience than charge you twice for the same month of work. No minimum and no commitment: switch to your own keys whenever you like and the charge stops. ### Which AI tools does it work with? Any tool that speaks MCP. Context and graph queries are served over a standard interface, so you are not locked to one assistant. ### What you get, and where it runs ### Is anything held back from a smaller plan? No. Every plan carries the whole platform. Plans differ only by how many repositories you hold and how much analysis you run each month. There is no feature flag separating them. ### Can we try it before paying? Yes, three ways. Install the CLI and run the engine on your own machine, which is free and needs no account. Start on the cloud on the free plan and connect a repository: 250 events a month, about 10 pull requests, for as long as you like. Move to a paid plan when you need more. ### Can we run it on-prem or air-gapped? Yes. On-premise deploys the full platform inside your network, premium plugins included, and supports air-gapped setups with your own model endpoints. There is no event metering on-prem. Talk to us at hello@sourceant.ai. Something not answered here? [Ask us directly](mailto:hello@sourceant.ai?subject=Pricing).