Anytime, in one click, from the Stripe billing portal in your dashboard. Your plan runs to the end of the paid period and stops — no partial-month refunds. And after signup you get free usage, so you can test Standard Compute and every claim before paying anything.

Current frontier modelsClaude Fable 5 · GPT-5.6 Sol

Your agents never stop.
Your bill never grows.

Frontier models when it counts. Efficient models when it doesn’t. No usage limits. One flat bill.

"Hi"GPT-5.6 Lunainexpensive model

Free trial No credit card needed Billing by Stripe3,500+ agent users+30.4% weekly growth ↗

Kudos guys! Really impressed with the model quality. Code output is better and not hitting usage limits feels fantastic! Went from about $200 a month to just using your Standard plan. You really set a new standard for what compute should look like.

Brandon Wu · customer email, July 25 · shared with permission

The new standard

Side by side.

	Pay as you go	Subscriptions	Standard Compute
No usage limits	✓	✕	✓
No “quota exceeded” errors	✓	✕	✓
No meter to watch	✕	✓	✓
More compute than you pay for	✕	~	✓
Smart routing	~	✕	✓
No 5-hour session caps	✓	✕	✓
Peace of mind	✕	✓	✓
A bill that never surprises	✕	✓	✓

01 · Smart routing

You save a ton of money.

Smart routing means no wasted tokens — and wasted tokens are wasted money.

2–7×more compute per dollaruser reported

$1.44saved vs frontier pricing150k context

"hi"Standard Compute

02 · No usage limits

Heavy loads are paced. Your agent never stops.

429 rate limitednever

“quota exceeded”never

“limit reached — try again in 5 hours”never

Under sustained heavy load, speed eases down instead of cutting you off — and recovers on its own. Your agent always runs.

The internet did it. Streaming did it. Compute is next.
Why Standard Compute exists

1996The internet~~paid by the minute~~ flat monthlySpeed depends on plan

2007Streaming~~paid per movie~~ flat monthlyResolution depends on plan

NowIntelligence~~paid per token~~ flat monthlySpeed depends on plan

03 · The economics

You get 2–7× more compute per $.

Three mechanisms stack in your favor.

04 · The models

Every frontier lab. Always current.

New models land within days, each one editorially reviewed — then served through the strongest providers, re-weighted in real time on price, speed and quality.^*

frontierClaude Opus 5Anthropic

frontierGPT-5.6 SolOpenAI

frontierGemini 3.1 ProGoogle

open frontierGLM 5.2Z.ai

open frontierKimi K3Moonshot AI

open frontierMiniMax M3MiniMax

* A representative lineup, not the full roster. Models and providers shift as we review and re-score them — new releases often land before this page catches up.

05 · Integration

Two minutes to switch.

Change the base URLapi.stdcmpt.com/v1
Paste your API keyfrom your dashboard
Set the model"standardcompute"

agent.config.ts+3 −3

const client = new OpenAI({
  baseURL: "https://api.openai.com/v1",  baseURL: "https://api.stdcmpt.com/v1",  apiKey: process.env.OPENAI_API_KEY,  apiKey: process.env.STANDARD_COMPUTE_KEY,});

await client.chat.completions.create({
  model: "gpt-5.2",  model: "standardcompute",  // the router takes it from here  messages,
});

# works for any agent that reads the standard env vars

export OPENAI_BASE_URL="https://api.stdcmpt.com/v1"
export OPENAI_API_KEY="sc_live_…"
export OPENAI_MODEL="standardcompute"

# restart the agent — done

curl https://api.stdcmpt.com/v1/chat/completions \
  -H "Authorization: Bearer $STANDARD_COMPUTE_KEY" \
  -d '{"model": "standardcompute", "messages": …}'

06 · Reliability

Boring, the way infrastructure should be.

api.stdcmpt.com — last 90 days99.97% uptime

90 days agolive status page →today

Live statusPublic pacing policyBilling by Stripe

Pricing

Pay for the lane, not the traffic.

One flat monthly price — speed is the only difference.

№ 01 / 03

Economy

$39/mo

Standard lane

No usage limits
Every model, smart-routed
Shared execution pool
Commercial use

Get Economy

Free trial included, no credit card

Common questions.

What's the catch?

Speed. Run far beyond your plan’s pooled share for a sustained period and responses slow down until usage rebalances — like traffic shaping on a network. You are never cut off and requests are never rejected. The exact policy is on the fair-use page, and your dashboard shows a live speed gauge, current pacing, and a recovery forecast.

How can a flat rate be sustainable?

The same way it works for your ISP. Most agent traffic doesn’t need a frontier model, so routing it to the right model cuts the real cost of a typical request by 10× or more. Plans then draw on a shared compute pool — quiet users and heavy users average out — and pacing keeps the extreme tail inside the pool. No loss-leader pricing, just the math every flat-rate utility runs on.

Will it work with my agent?

If it speaks the OpenAI API — and nearly every agent does — yes. Three edits: point the base URL at api.stdcmpt.com/v1, paste your key, set the model to "standardcompute". Typical setup is under two minutes; step-by-step guides for every major agent.

Is my data used to train models?

Not by us — your prompts and outputs are never stored, logged, or used to train Standard Compute’s models. Upstream, training policies vary by provider, so provider selection is security-first: we review every provider’s data handling before routing to it, and drop any that fall short. Details on the security page.

Which models do I get?

Frontier models from every major lab — Claude Opus 5, GPT-5.6 Sol, Gemini 3.1 Pro — alongside open-frontier models like GLM 5.2, Kimi K3, and MiniMax M3 for routine traffic. The router picks per request, and new releases are reviewed and added within days.

Can I cancel?

Anytime, in one click, from the Stripe billing portal in your dashboard. Your plan runs to the end of the paid period and stops — no partial-month refunds, no questions asked. And after signup you get free usage, so you can test Standard Compute — and every claim on this page — before paying anything.