Xpersona / 01OpenAI-compatible inference

One key.Every frontier model.Ship the work.

Use genuine GPT, Claude, and Gemini models alongside Xpersona originals through one OpenAI-compatible API, one dashboard, and one bill. Build with real GPT, Claude, Gemini, and Xpersona originals through one key across chat, OpenCode, and any OpenAI-compatible client — with a usage dashboard that keeps every request and every dollar in view.

  • GPT
  • Claude
  • Gemini
  • Xpersona
  • Council
Try a run now

Builder from $20 / month · prepaid credits from $2 · no automatic refill

A dreamlike Hivemind figure surrounded by colorful winged creatures
MODEL / 01CAPACITY THAT COMPOUNDS
ChatOpenCodeAPIUsageTrust
One keyDrives chat, OpenCode, and the API from a single account.
60–70% below listOn many premium frontier models compared to official pricing.
Live usageEvery request lands in a dashboard you can actually read.
Trust, shownBenchmarks, reliability, and methodology published on-site.

The work / in motion

One subscription.
Code anywhere.

Move from model choice to working code inside the Xpersona desktop experience, with one account and the same catalog behind every surface.

XPERSONA / DESKTOP 01MODEL · CODE · SHIP

The Xpersona field guide

One system.
Many minds.

Genuine modelsOne keyVisible usage

01 / Unify

One key / every model

XPERSONA / ARCHIVE 01

Use genuine GPT, Claude, Gemini, and Xpersona models through one OpenAI-compatible account.

02 / Reason

Match the mind / to the work

XPERSONA / ARCHIVE 02

Move between frontier coding, deep analysis, fast multimodal work, and Xpersona originals without changing clients.

03 / Choose

Capacity / on your terms

XPERSONA / ARCHIVE 03

Builder, Pro, and Studio make included monthly capacity visible before you start. Prepaid credits begin at $2.

04 / Delegate

Let the agent / carry the load

XPERSONA / ARCHIVE 04

Give coding work a consistent endpoint while your tools stay focused on the problem at hand.

05 / Verify

Proof / before promises

XPERSONA / ARCHIVE 05

Published benchmark results, live model routes, and methodology notes let you inspect the record before you buy.

06 / Observe

Usage / stays in view

XPERSONA / ARCHIVE 06

Requests, tokens, package state, and successful usage stay together in the same Xpersona dashboard.

Predictable monthly capacity

Choose your
working room.

Every monthly package unlocks the complete public catalog. Premium models use capacity faster than lighter models, and included capacity pauses without surprise overages.

Monthly / Builder

Builder

$20/ month
Included each billing month1x monthly capacity

A focused lane for individual builders and everyday agent work.

  • Full model catalog
  • Supplier-weighted capacity
  • No overage billing

Monthly / Studio

Studio

$200/ month
Included each billing month20x monthly capacity

The highest public allowance for teams and sustained heavy runs.

  • Everything in Pro
  • 20x Builder capacity
  • Highest public allowance

Prefer variable spend?

Top up from $2 with no subscription or automatic refill.
Compare every option

The short way through

Four moves.
No maze.

One account from model choice to finished work, with setup and usage kept in view.

Open setup guide
01

Choose the model

Match GPT, Claude, Gemini, or an Xpersona original to the work in front of you.

02

Connect one key

Use the same Xpersona account in OpenCode, chat, or another OpenAI-compatible client.

03

Ship the work

Keep your agent in flow with one familiar endpoint and a consistent model catalog.

04

Read the receipt

Review requests, tokens, package capacity, and successful usage from the dashboard.

Xpersona / live workspace

SYSTEM READY
MODELSUSAGERECEIPTS

$ xp models --available

OpenAI    GPT-5.6 Sol

Anthropic Claude Opus 4.8

Google    Gemini 3.5 Flash

Xpersona  Frieren 1

✓ public catalog ready · one API key

SELECTED FOR AGENT WORKClaude Fable 5extra-high reasoning · OpenCode ready
Xpersona // muse_01signal online
Monika // source portrait ASCII74 x 64 glyph signal

Xpersona Council · live comparison

Ask two models. Compare the evidence. Continue with the winner.

Run the same task against real Xpersona models through the live runtime. In Council you add a blind judge and synthesize; here is the raw comparison, no account required.

Hit "Compare models" to run this exact task through both Frieren 1 and A_BETTER_YOU right now.

The highlighted model returned the most complete answer for this run. Council formalizes this with a blind, rubric-based judge and a shareable receipt.
Explore Council

Proof before promises

Inspect the record before you buy.

Benchmarked results, live reliability, and a documented method are part of the product.

Share Xpersona

Get other builders on one key.

Send your personal referral link. Signups through your link are attributed to you, so the people you bring ship on the same models — and your referral is easy to trace.

https://www.xpersona.co/?ref=xp_psjpwv&utm_source=xpersona-ref&utm_medium=organic&utm_campaign=referral&utm_content=share_card

Share to your community

Attribution uses the ref and utm_source=xpersona-ref params, so signups from your link are tracked through the funnel on your domain.

WeChat users: copy the link above and forward it to a group or Moments. Your referral is tracked on your own domain.

Your model lineup is ready

Give your agent
room to think.

Most premium models are 60-70% below standard official API list prices. Start with Builder, connect one key, and choose the right model for every request.