Pricing
Pick your runs. AI at cost.
Every plan includes your agent's own cloud computer and unlimited interactive sessions. Plans differ only in how many workflow runs you get each month. AI usage is passed through at exactly what our providers charge us.
Base
$10/ month
750 workflow runs / month
≈ one hourly workflow
For a few automations on a schedule, plus everyday agent sessions.
Pro
$30/ month
3,000 workflow runs / month
≈ four hourly workflows
For a handful of hourly workflows and webhooks running around the clock.
Max
$100/ month
Unlimited workflow runs
For agents that never stop. No run limit.
Every plan
Provisioning and management for your agent's personal sandbox. The free trial runs with base limits.
- Your own personal cloud sandbox, provisioned and managed for you
- MCP servers, skills, tools, and files synced with your laptop
- Schedule, webhook, and hosted-form triggers
- Session history and Optimize analytics
- Unlimited interactive sessions in every plan
AI usage
At cost
Standard token pricing from our providers. You pay for what your agents use.
- Every model at exactly the provider's price, no markup
- Charged from a prepaid wallet when each agent turn finishes
- $10 top-ups, with optional auto-recharge below $2
- No API keys to manage, locally or in the cloud
Model prices
Input and output prices in USD per 1M tokens, the same price our providers charge us.
Text models
| Model | Context | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
| anthropic/claude-opus-5.5 | 1M | $4.00 | $20.00 |
| anthropic/claude-sonnet-5.5 | 1M | $2.00 | $10.00 |
| deepseek/deepseek-v4.1-flash | 1M | $0.040 | $0.30 |
| google/gemini-4-argon | 1M | $2.00 | $10.00 |
| meta/muse-ember | 1M | $0.10 | $0.50 |
| mistral/mistral-large-4 | 1M | $0.70 | $2.10 |
| moonshotai/kimi-k3 | 1M | $0.16 | $10.00 |
| openai/gpt-6-luna | 1.10M | $0.10 | $0.50 |
| openai/gpt-6.1-sol | 1.10M | $2.00 | $10.00 |
| openai/gpt-6.1-terra | 1.10M | $1.00 | $5.00 |
| particle/particle-1 | 1M | $0.13 | $0.50 |
| reflection/beam | 1M | $0.25 | $1.00 |
| zai/glm-5.3 | 1M | $0.070 | $1.40 |
| zai/glm-5.3-flash | 1M | $0.040 | $0.080 |
Decision models
| Model | Context | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
| google/gemini-decision-1 | 262k | $0.10 | Free |
| liquid/d1 | 66k | $0.020 | Free |
| microsoft/microsoft-decision-1 | 33k | $0.040 | Free |
| openai/gpt-6-luna-decisions | 1.10M | $0.10 | Free |
| perplexity/pplx-decider-v1.2 | 262k | $0.020 | Free |
| typesafe-ai/jev | 32k | $0.040 | Free |
What is the sandbox?
A personal cloud computer for your agent, with its own filesystem and shell. Your MCP servers, skills, and tools sync to it, so workflows run there the same way they run on your laptop.
What happens when my wallet runs low?
Turn on auto-recharge and we add $10 whenever your balance drops below $2. Otherwise, top up from the billing page whenever you like.
What happens if I hit my run limit?
Your workflows pause: scheduled, webhook, and form runs don't start until you upgrade or your runs reset at the start of your next billing period. We email you and show a notice, and nothing about your workflows or triggers changes.
Can I only run fl0w locally?
Yes. The CLI is open source. Model calls still go through your fl0w account, so AI usage is billed at cost from your wallet.