Stop wasting tokens

Save tokens
with reasoning traces.

Your AI saves what it already worked out and reuses it — so it burns far fewer tokens on every task. Host it for your whole team in a few clicks, or run it free on your own machine.

Few-click hosted setup for teams — or free on your own computer, no account needed.

How you save

Fewer tokens on every task.

Your AI reuses what it already worked out, instead of spending tokens to generate it again from scratch.

Reuse, don't regenerate

Your AI reads what it already worked out — instead of spending tokens to figure the same thing out all over again.

Long tasks stay lean

Big jobs usually burn tokens repeating themselves. Reusing earlier context keeps the token count low from start to finish.

See tokens saved across your team

Everyone's reused context adds up. Your dashboard shows the tokens — and the spend — you're saving company-wide.

No new workflow

Works with the AI tools you already use.

Point Claude, Codex, or Cursor at Scholialang. Nothing about how you work changes — your AI just reuses context and spends far fewer tokens.

~/project
# without traces — full token cost every run
$ run task# 1,240,000 tokens
$ run task# 1,240,000 tokens again…
# with traces — reuses earlier context
$ run task --resume# ~180,000 tokens
Simple pricing

Cloud plans for your team.

Hosted plans differ by how much you run, whether you're a team, and whether the cloud writes your traces for you. Prefer to self-host? Local-first stays free, forever.

Starter

Your traces, off your laptop.

$5/person/mo

Solo, just moving to the cloud.

  • Hosted trace storage
  • Everyday usage limits
  • 7-day retention
Start 14-day trial →

Pro

For heavy solo use.

$12/person/mo

Individuals running lots of traces.

  • Higher usage limits
  • Official app relay
  • 30-day retention
  • Gemma4 Cloud — pay-as-you-go
Start 14-day trial →

Business

Teams + AI writing, offloaded.

$24/person/mo

Teams that need collaboration, admin & compliance.

  • Everything in Pro
  • Gemma4 Cloud included
  • SSO, SCIM & audit log
  • Admin & configurable retention
Start 14-day trial →
Gemma4 Cloud

Let the cloud write your traces.

Normally your machine runs the Gemma model to write and summarize every reasoning trace. On the upper plans a hosted Gemma4 model does it in our cloud — no local compute, faster, and consistent. Included on Business, pay-as-you-go on Pro.

Compare plansFreelocalStarter$5/moPro$12/moBusiness$24/mo
Local-first (SQLite, exports)✓✓✓✓
Hosted cloud trace storage—✓✓✓
Usage limits—BasicHigherHighest
Search across traces—✓✓✓
Official app relay——✓✓
Retention windowLocal7 days30 daysConfigurable
Gemma4 Cloud — offloaded trace writing——Pay-as-you-goIncluded
Team workspaces & shared projects———✓
Members, roles & seat admin———✓
SSO (SAML / OIDC) & SCIM———✓
Audit log & retention controls———✓
Dedicated storage / BYO-Cloudflare———✓
SupportCommunityCommunityEmailPriority

Hosted trace writes via MCP

Every reasoning trace your agent writes to the cloud is one MCP trace write. The monthly allowance is set by your workspace tier and pooled across every seat — beyond it, writes are metered at $0.05 per 1,000.

Free local-firstLocal onlyUnlimited on your own machine · nothing hosted
Team workspace200,000/mo$19/mo workspace · pooled across seats
Enterprise workspace20,000,000/mo$750/mo workspace · pooled across seats
Running at enterprise scale?20M trace writes/mo, SSO/SCIM, audit log, and dedicated storage on an Enterprise workspace.
Start Enterprise trial →
MCP requests$0.20 / 1,000 MCP requests
Trace writes$0.05 / 1,000 trace writes
Search scans$0.01 / 1,000 search scans

The free plan keeps 1user on your own computer — it isn't a trial that runs out.

Spend fewer tokens on every task.

Get your team on hosted reasoning traces — or run it free on your own machine.