Frontier models,
from $10 a month.

No. Starter is free with no payment information required. Sign in on web, desktop, or CLI and start working.

Early Access pricing for Pro ($10/mo) is available for a limited time, and we'll give notice before it closes to new subscribers. If you subscribe during Early Access, that $10/mo rate continues on each renewal for as long as your subscription stays active, and we change the price you pay only with advance notice. After Early Access ends, new subscribers pay the $20/mo list price.

Starter includes 20,000 PT a month. Pro has two separate hosted-usage lanes: Roost has no fixed cap, with fair-use limits over 5 hours and 7 days; manual model selection allows 15,000 PT every 5 hours, 37,500 PT per rolling 7 days, and 75,000 PT per billing period. The 5-hour limits reset every 5 hours, and the weekly limits roll. Reaching a limit in one Pro lane does not pause the other. Ultra has no 5-hour limit on either lane: Roost has no fixed cap, with a weekly fair-use limit, and manual model selection allows 112,500 PT per rolling 7 days and 225,000 PT per billing period.

No 5-hour limit, three times Pro's weekly limits on both usage lanes and three times its billing-period allowance for manual model selection, plus Claude Fable 5.2 and GPT-6 Astra. Roost has no fixed cap, with a weekly fair-use limit. Manual model selection allows 112,500 PT per rolling 7 days and 225,000 PT per billing period. A notice appears at 50% and 80% of each weekly limit. Everything else matches Pro.

Yes. Moving from Pro to Ultra keeps your subscription and billing date: the prorated difference for the rest of the current billing period is charged right away, and Ultra's limits apply as soon as that payment succeeds. To move back, choose Switch to Pro at renewal in Settings → Billing; Pro starts at your next renewal at its $10 price, and until then you can choose Keep Ultra to stay.

Claude Fable 5.2's host keeps prompts and replies for up to 30 days and may have people review them for safety. Security and data explains how Perch handles model requests. Ultra activity is excluded from the Starter Data Program, like Pro.

Starter's free lineup includes Laguna S 2.1, DeepSeek V4.1 Flash, and GLM 5.3 Flash, plus Qwen 3.8 Max until November 20, 2026. Usage still counts toward Starter's included allowance. GPT-6.1 Sol is Pro only. Pro includes the full Starter lineup and every premium model except Claude Fable 5.2 and GPT-6 Astra, led by Gemini 4 Argon, then Claude Opus 5.5, GPT-6.1 Sol, GPT-6 Luna, Claude Sonnet 5.5 and Claude Haiku 5.5. Claude Fable 5.2 and GPT-6 Astra are on Ultra. See the Models guide for current availability and the Model promotions guide for how promotions work.

Yes, on every plan including Starter. Some of the strongest models are text-only and cannot open an image at all. Perch reads the image for them, so you can share a screenshot, a scanned invoice, or a photo with any model and still get an answer. There is nothing to enable, and the reading step does not count against your included monthly usage.

Yes, on every plan including the free one. Add a key from any OpenAI-compatible provider, or point Perch at your own endpoint, and name any model you want. Perch usage limits do not apply to inference you pay for yourself, and your included allowance stays untouched while you use it.

Pro, Ultra and Enterprise activity, including complimentary Pro or Ultra access, is excluded from training and dataset licensing. Bring-your-own-key and self-hosted usage is excluded on every plan. Eligible Perch-funded Starter usage may contribute to Derived Training Data for AI training, evaluation, and benchmarking after aggressive deidentification and transformation; raw or reasonably linkable content is never licensed. The Starter Data Program is enabled by default after notice and can be turned off any time in Settings → Privacy.

No. Starter, Pro and Ultra include Perch-hosted model usage through Perch AI Web, Desktop, and CLI. Perch's internal application endpoints are not a public API and direct calls may be blocked or rate limited. Repeated attempts may result in account suspension. Contact us if you need API access for an integration. Bring-your-own-key usage inside Perch is still available on every plan.