# Plansmith Pricing: Priced per planner. Paid back in weeks.

> Plansmith is the planning agent. It turns business requirements into build-ready specs and prototypes and pushes them to Linear, Jira, Asana, or Radial in one click. Every plan includes the whole product; tiers differ by usage and by what a team needs.

One seat for each person who plans. Your developers never need one, and the AI runs at cost.

Pricing is per user, billed monthly or yearly. There is no free tier and no trial. To see it work before you buy, book a demo and bring real business requirements from your backlog.

## The return: a worked example

13 planner hours per build-ready spec and prototype down to 3 (writing 8 to 2, questions 5 to 1), so each planner writes build-ready specs and prototypes 4.3× faster. A team of 100 engineers and 40 planners gets back $4.4M of team time a year, valued at salary (time, not cash). That is a 35× return on investment, paid back within 2 weeks.

| Where it comes from | Assumption | Per year |
| --- | --- | --- |
| Rework | 30% of engineering capacity, cut by 40% | $2.2M |
| Questions back and forth | 10.0 hrs down to 2.0 hrs per spec | $1.3M |
| Writing specs & prototyping | 8 hrs down to 2 hrs per spec | $969K |
| Wasted agent runs | 30% of 100 agent seats wasted, 50% of it recovered | $36K |
| Plansmith seats | 60 full seats, 80 comment-only, billed yearly | −$124K |
| **Saved each year** | | **$4.4M** |

Conservative case: $1.9M a year after seats, a 15× return and 2.6× planner output, with specs at 8 to 4 hours and rework 8% of capacity cut by 15%.

An illustrative model, not a customer result. Time is valued at fully loaded US salaries over 2,080 hours; wasted agent runs are cash. Clarification and rework may overlap. Excludes top-ups and rollout time. Run it with your own numbers: https://www.plansmith.co/roi. The business case as a PDF: https://www.plansmith.co/roi/plansmith-the-cost-of-poor-planning.pdf.

## Your AI budget, under control.

Every seat includes a monthly budget of model usage, billed at cost. There are no overage bills: when a budget runs out, planning pauses until you add more.

- **See every planner’s usage.** One page shows what each person has used this month, and flags anyone running low.
- **Move usage to your heavy users.** Take it from someone who isn’t using theirs and give it to someone who is. You see what each person keeps before you confirm, and the reason lands in the audit log.
- **Add usage when you need it.** Top up with $50 or $200 at any time, for one person or the whole team.

## The AI at cost

Model usage is billed at the models’ published price, with nothing added. Solo is 68% model usage and 32% platform; Team is 85% and 15%.

| Plan | Price per user, per month | Model usage, at cost | Platform |
| --- | --- | --- | --- |
| Solo | $50 | $34 (68%) | $16 (32%) |
| Team | $200 | $170 (85%) | $30 (15%) |

Nothing is added to the model price, and card fees come out of the platform part. Yearly billing takes 2 months off the price on top.

Need more? Add $50 or $200 of usage anytime. Each splits exactly like the plan at the same price.

## Plans at a glance

| Plan | Billed monthly | Billed yearly | Yearly, per month | Model usage per user, per month, at cost | Platform share, billed monthly |
| --- | --- | --- | --- | --- | --- |
| Solo | $50 per user | $500 per user | $42 | $34 | 32% |
| Team | $200 per user | $2,000 per user | $167 | $170 | 15% |
| Enterprise | Contact us | Contact us | Contact us | Sized to your team | Lower, by volume |

## Who needs a seat

| Who | What they do | Seat | Cost |
| --- | --- | --- | --- |
| Planners | Create and refine plans with the agent. | Full seat | Solo or Team price |
| Teammates | Read everything and comment inside Plansmith. Comments do not run the agent, so this seat includes no model usage. | Comment-only seat | $5 per user, per month |
| Developers | Work from the issues Plansmith pushes to Linear, Jira, Asana, or Radial. | No seat | $0 |
| Stakeholders | Review a shared prototype and comment through a public link. | No seat | $0 |

## Solo

$50 per user, per month, or $500 per user, per year. For planning on your own.

- $34 of model usage a month, at cost (68% model usage · 32% platform).
- One user
- Interviews that end in build-ready specs and prototypes
- Your code and your business context
- Prototypes you can click and share
- One-click push to Linear, Jira, Asana, or Radial

## Team

$200 per user, per month, or $2,000 per user, per year. For planning with other people.

- $170 of model usage a month, at cost (85% model usage · 15% platform).
- Everything in Solo, for your whole team
- Comments and sharing, so a plan can be reviewed
- Comment-only seats at $5 for everyone else
- Single sign-on
- Audit logs: who changed what, and when
- Self-hosted GitHub, GitLab, or Bitbucket
- Priority support: your questions go to the front of the queue

## Enterprise

Contact us. For the organization that plans at scale.

- Model usage at cost, sized to your team. A lower platform share, set by volume.
- Everything in Team
- Volume pricing for large groups
- Dedicated onboarding: we set up your workspace and Context with you
- Custom contracts, invoicing, and procurement support
- A named contact for support

## What the models cost

Each maker's published price per million tokens. Your model usage is billed at exactly these rates: this table is generated from the same rate card that meters your usage.

Prices in US dollars per million tokens.

| Model | Maker | Input | Cached input | Output |
| --- | --- | --- | --- | --- |
| Sonnet 5.5 | Anthropic | $2.00 | $0.20 | $10.00 |
| Opus 5.5 | Anthropic | $4.00 | $0.20 | $20.00 |
| Haiku 4.5 | Anthropic | $1.00 | $0.10 | $5.00 |
| GPT-6 Astra | OpenAI | $10.00 | $1.00 | $50.00 |
| GPT-6 Sol | OpenAI | $2.00 | $0.20 | $10.00 |
| GPT-5.6 Terra | OpenAI | $2.00 | $0.20 | $12.00 |
| GPT-6 Luna | OpenAI | $0.10 | $0.01 | $0.50 |
| Gemini 3.1 Pro * | Google | $2.00 | $0.20 | $12.00 |
| Gemini 3.6 Flash | Google | $1.50 | $0.15 | $7.50 |
| Gemini 3.8 Flash | Google | $0.75 | $0.075 | $3.75 |
| Grok 4.7 * | xAI | $1.60 | $0.40 | $4.80 |

Marked models charge a higher long-context price for prompts over 200,000 tokens. Cached input is the price of re-reading a prompt the model has already seen, which is most of a planning conversation. Writing to that cache is billed at the maker's cache-write price.

### Other work the agent pays for

- Web search: $0.007 per search
- Reading a web page: $0.001 per page
- Deep research: $0.10 to $1.00 per research provider, by depth
- Transcribing a recording: $0.40 per audio hour

## What each plan includes

| Capability | Solo | Team | Enterprise |
| --- | --- | --- | --- |
| Model usage per user, per month, at cost | $34 | $170 | Sized to your team |
| Platform share of the price, billed monthly | 32% | 15% | Lower, by volume |
| Comment-only seats | $5 per user/month | $5 per user/month | $5 per user/month |
| Interviews, build-ready specs and prototypes | Yes | Yes | Yes |
| Readiness and effort scores | Yes | Yes | Yes |
| Numbered acceptance criteria and build order | Yes | Yes | Yes |
| Code context: GitHub, GitLab, Bitbucket (read only) | Yes | Yes | Yes |
| Business context: documents, links, MCP sources | Yes | Yes | Yes |
| Company, project, and plan levels | Yes | Yes | Yes |
| Clickable prototypes | Yes | Yes | Yes |
| Public share links and pinned comments | Yes | Yes | Yes |
| Invite your whole team | Yes | Yes | Yes |
| Comments and live changes | Yes | Yes | Yes |
| Planner and read-only roles | Yes | Yes | Yes |
| One-click push: Linear, Jira, Asana, Radial | Yes | Yes | Yes |
| Markdown and CSV export | Yes | Yes | Yes |
| Single sign-on | No | Yes | Yes |
| Audit logs | No | Yes | Yes |
| Self-hosted GitHub, GitLab, Bitbucket | No | Yes | Yes |
| Support | Standard | Priority | Named contact |
| Dedicated onboarding | No | No | Yes |
| Volume pricing, custom contracts, procurement | No | No | Yes |

## Questions

**How does usage work?**

Every time the agent reads your code, drafts a spec, or builds a prototype, the AI provider bills us for the tokens it uses. That is model usage, and you pay it at cost: the provider's own price, with nothing added. The rest of your price pays for the platform. A Solo seat is $34 of model usage and $16 of platform each month; a Team seat is $170 and $30. Usage pools across your team.

**Why is the platform share lower on Team?**

A bigger commitment carries a smaller share: 32% of the price on Solo, 15% on Team. If you measure it against the model cost instead of the price, the platform adds 47% on Solo and 18% on Team. Same numbers, two ways to read them, and we would rather you see both.

**Does unused usage roll over?**

No. Monthly usage resets on your billing date. Usage you add on top lasts 12 months.

**What happens if I run out?**

Planning pauses, and nothing is lost. Add $50 or $200 of usage anytime, or move up a plan. A $50 top-up is $34 of model usage; a $200 top-up is $170, on any plan.

**Which plan should I pick?**

Solo, if you plan on your own. Team, the moment a second person is involved: multi-user, comments and sharing, five times the model usage per planner, and a smaller platform share. Enterprise, when you need volume pricing or custom terms. Upgrade anytime; your Projects, Plans, Specs, and Context all come with you.

**Who actually needs a seat?**

Planners need full seats: the people who create and refine plans. Teammates who follow along inside Plansmith take comment-only seats at $5 per user, per month: they read everything and can comment, but not edit. Your developers work from the issues Plansmith pushes to the issue tracker, and prototype reviewers comment through a public link; neither needs a seat at all.

**Is there a free tier?**

No. If you want to see it work before you buy, book a demo and bring real business requirements from your backlog. We plan them live.

**Can't I do this with a chatbot?**

A chat window gives you text you still have to turn into tickets by hand. Plansmith reads your team's code and documents, keeps the whole plan of Specs, scores each one, and files the issues in your issue tracker itself. Read access only: the agent never changes your code. The chat is where planning starts, not what you get.

**Do we need an issue tracker to start?**

No. You can plan, prototype, and export from day one: any plan becomes a plain document (Markdown) or a spreadsheet (CSV). Connect Linear, Jira, Asana, or Radial when you're ready to push.

**What makes Team worth four times Solo?**

People, mostly. Solo is one user; Team is everyone, with comments and sharing so a plan can be reviewed. It also gives each planner five times the model usage at a smaller platform share, plus priority support. Weigh it against what one bounced ticket costs: a grooming meeting, a rewrite, a week of back and forth.

## Links

- Pricing page: https://www.plansmith.co/pricing
- Connectors: https://www.plansmith.co/connectors
- Book a demo: https://www.plansmith.co/book-a-demo
- Index for agents: https://www.plansmith.co/llms.txt