Imbue pricing

About

Imbue is a San Francisco AI research lab building models and tools for robust reasoning and coding agents. It launched publicly as Generally Intelligent in October 2022 and rebranded to Imbue in September 2023, when it announced a reported $200 million Series B at a valuation over 1B USD — vaulting it to unicorn status. Reported total funding sits around 220M USD, with investors including the Astera Institute, NVIDIA, Cruise CEO Kyle Vogt, and Notion co-founder Simon Last. The lab is led by co-founders Kanjun Qiu (CEO) and Josh Albrecht (CTO).

Here is the thing you need to understand before you go looking for a pricing page: Imbue does not sell a commercial product. It is research-stage. At the time of its raise it said plainly that it was not commercializing — it was building foundational reasoning models (trained on a cluster reported at 10,000 NVIDIA H100 GPUs) toward a future platform for building custom AI agents. The products it has actually shipped to users are free, developer-facing tools — most prominently Sculptor, a UI for running parallel coding agents, alongside open-source utilities like mngr, Bouncer, Blueprint, Vet, and Latchkey.

So when you ask “what does Imbue cost?”, the honest answer is: nothing, today. The cost you should be modeling is your Anthropic bill, because Sculptor is bring-your-own-key (BYOK) — it drives Claude Code using your own subscription or API key.

For the most current information, visit Imbue.

Pricing summary : How Imbue’s pricing model works

Imbue has no public rate card and no commercial pricing model. This is not a gated-enterprise situation where a number exists behind a sales call — it is a research lab that has not monetized. Its flagship shipping tool, Sculptor, is free to use while in beta, distributed as a free download for Mac (Apple Silicon) and Linux. There is no checkout, no paid tier, and no waitlist toll.

The mechanic that actually matters for your wallet is BYOK (bring your own key). Sculptor orchestrates Claude Code agents but does not resell model access. You supply your own Claude Code subscription (Claude Pro/Max) or your own Anthropic API key, and all token cost flows to Anthropic on your existing bill. Imbue takes no fee and adds no markup. That means your spend is governed entirely by Anthropic’s metered token pricing and how many parallel agents you run — not by anything Imbue charges.

What makes this different: most tools in the AI-coding space wrap a model and add a per-seat or per-credit margin on top. Imbue’s shipping posture is the opposite — a free orchestration layer with zero take-rate, monetization deferred entirely to a future commercial platform that does not yet have a price.

Pricing by product

Sales motions across products: there is effectively no commercial sales motion today — the tools are simply free and self-distributed. We classify the eventual commercial track as sales-led because any future enterprise platform from a frontier lab would almost certainly be quoted, not self-serve checkout. No real numbers exist to publish.

Hidden costs : What Imbue users actually pay

The “hidden cost” with Imbue is not hidden at all — it is your Anthropic bill. Sculptor is free, but it does real work by calling Claude Code, and those tokens are billed to your own Anthropic account. Run many agents in parallel and your Anthropic usage scales accordingly. Budget for the model, not for Imbue.

Want to estimate your model-side spend when driving agents? Use the Imbue pricing calculator to model the Anthropic token cost behind your agent runs.

Pricing evolution : Imbue pricing history and changes

Cadence

Tracked range: 2022–present. Imbue has never published a price for a commercial product; there is no Wayback pricing page to track because none has existed.

Notable changes

• Oct 2022 — Launches publicly as Generally Intelligent, a foundational-research lab.
• Sept 2023 — Rebrands to Imbue; announces reported $200M Series B at $1B-plus valuation. States it is not commercializing a product.
• Sept 2025 — Releases Sculptor as a free beta (free download, BYOK on your Anthropic access).
• June 2026 — Still no rate card; shipping tools remain free, future commercial pricing would be sales-quoted.

What’s unique : Imbue’s distinctive pricing mechanics

1. A $220M-funded company that charges users $0. Most heavily funded AI labs ship a priced API or seat-based product to start generating revenue. Imbue has deliberately deferred monetization — its only widely used tool is free, which is unusual for a company at its funding scale.

2. Zero take-rate BYOK. Sculptor wraps Claude Code but adds no margin. Many “AI coding workspace” competitors mark up tokens or charge a per-seat fee on top of the underlying model. Imbue’s deliberate non-pricing means the user keeps 100% of the model-cost relationship with Anthropic.

3. Open-source as distribution, not monetization. Tools like mngr, Bouncer and Vet are released openly with no commercial gate — distribution and goodwill ahead of any revenue model, consistent with the lab’s “honest software in your control” philosophy.

Strengths & weaknesses

Billing UX : Imbue billing controls and transparency

• Billing controls — None from Imbue, because Imbue does not bill you. Your spend controls live in your Anthropic account (API limits, plan caps).
• Usage visibility — Sculptor runs locally and you control what data it sends back to Imbue; token usage shows up on your Anthropic dashboard, not an Imbue invoice.
• Payment options — Not applicable. There is no Imbue checkout. Download Sculptor free and connect your existing Anthropic key or subscription.

Strategic wins : Why Imbue’s pricing decisions worked

1. Free + BYOK lowers adoption friction to zero

By charging nothing and riding the developer’s existing Anthropic relationship, Imbue removes both the price objection and the procurement step. For a research lab seeking real-world agent usage data, free distribution buys exactly what it needs: users. See usage-based pricing strategy for why metering can be deferred when the goal is adoption, not revenue.

2. No markup builds trust in a skeptical developer audience

Refusing to take a cut of model spend is itself a positioning statement. It differentiates Imbue from tools that quietly inflate token costs. Related: how AI companies structure pricing and outcome-based pricing trends.

3. Deferring monetization protects optionality

With substantial capital in the bank, Imbue can wait to price a future platform until it knows what the durable value metric is — rather than locking into a per-seat or per-token model prematurely. See choosing the right usage metric.

Areas to improve : Gaps in Imbue’s pricing approach

1. No path for teams that want to pay

There is no way for an enterprise to buy support, an SLA, or a contracted commitment. Teams that would happily pay for reliability have nowhere to spend. See bill shock and cost unpredictability — here the unpredictability is on the vendor side: when (and how) Sculptor monetizes is unknown.

2. Free-beta ambiguity

“Free while in beta” tells you nothing about the eventual price. Adopters building workflows on Sculptor are exposed to an undefined future cost. Clearer signaling of intended commercial direction would help teams plan.

3. No transparency on the future model

Because nothing is published, buyers can’t compare Imbue to priced competitors on a like-for-like basis. A directional statement (“we’ll be per-seat” or “we’ll stay BYOK and monetize a hosted tier”) would let the market plan around it.

Key takeaways

1. There is no Imbue rate card. It is a research lab; its tools are free, so price_transparency is sales-only by default — nothing is published.
2. Sculptor is free in beta and BYOK. You pay Anthropic, not Imbue — model your Anthropic token cost, not a SaaS fee.
3. Zero take-rate is the differentiator. Imbue adds no markup on model usage, unusual among AI-coding tools.
4. Funding decouples product from revenue. A reported $220M raise lets Imbue defer monetization and keep tools free.
5. Future pricing is unknown. Any commercial platform would be sales-quoted; budget for change if you build on the free beta.

UBP implications

1. Free + BYOK is a valid go-to-market for a research lab — let users absorb the metered model cost directly while you gather usage data, then layer pricing later.
2. Zero markup can be a trust play. When the underlying model is the cost driver, refusing to mark it up differentiates you from wrappers that inflate it.
3. Deferring the value metric is a feature, not a gap, early on — Imbue avoids locking into per-seat or per-token before it knows what customers will pay for.

Sources

• Imbue official website (accessed 2026-06-16)
• Sculptor product page — free beta, BYOK (accessed 2026-06-16)
• Imbue (@imbue_ai) on X — “Sculptor is free to use while in beta…” (accessed 2026-06-16)
• TechCrunch: Imbue raises $200M to build AI models that can robustly reason (published 2023-09-07, accessed 2026-06-16)
• Bloomberg: AI Startup Imbue Tops $1 Billion Valuation After Funding from Nvidia (published 2023-09-07, accessed 2026-06-16)
• Crunchbase News: New AI unicorn Imbue (Astera, Nvidia) (accessed 2026-06-16)

Bottom line

Imbue is a well-capitalized reasoning-agent research lab (formerly Generally Intelligent), not a priced SaaS company. It has no public rate card because it sells nothing commercial today. Its flagship tool, Sculptor, is free in beta and BYOK — your only cost is your own Anthropic bill, and Imbue takes no markup. If you’re evaluating it, budget for Claude tokens, not an Imbue invoice, and plan for the possibility that a commercial tier (sales-quoted) arrives later. Browse the pricing blueprint for fully-researched company profiles.

Want to compare Imbue against other AI-coding and reasoning-agent companies? Browse the pricing blueprint.