Live across 15 providers · 16 models

The inference
price layer.

Real-time AI inference pricing across every major AI provider. One API. Every provider. Stop overpaying for inference.

See live prices →

Live Prices · $/1M blended

How it works

One call. Every provider.

We normalize pricing from 15 providers into a single, queryable schema so you can stop doing it yourself.

⚡

Real-time fetch

Every request hits provider APIs directly. Results are normalized to $/1M and cached for 60 seconds.

📊

OHLC price history

Every price event is captured and aggregated into daily open/high/low/close candles — so you can see if prices are trending up or down.

🔔

Price alerts Developer

Set a threshold and get notified via email or Slack the moment any provider crosses it. Never overpay again.

🔑

One clean API

Your cxm_ key, one base URL, three endpoints. Drop it into your cost pipeline or monitoring stack in minutes.

Developer-first

A price API that
actually makes sense.

Three endpoints. Consistent schema. CORS-open for browser use.

/prices — current prices across all providers
/compare — ranked list with cheapest pick + spread
/history — OHLC per provider per model (7-day free · 30-day paid)
Input, output, and blended metrics on every response
X-Coaxiom-Source header tells you if it's live or cached
Hard 429 with retry-after — never silent failures

# Get cheapest Llama 3.1 70B provider right nowcurl https://coaxiom.io/api/v1/compare?model=llama-3.1-70b \
  -H "Authorization: Bearer cxm_..."

// Response
{
  "model": "llama-3.1-70b",
  "cheapest": {
    "provider": "Groq",
    "blended_per_1m": 0.59
  },
  "spread_pct": 340,
  "spread_usd": 1.41,
  "recommendation": "Use Groq — 340% cheaper than most expensive",
  "ranked": [...]
}

Coming soon

Got spare inference capacity? Sell it.

List any OpenAI-compatible endpoint on the Coaxiom exchange. Set your price. We proxy the calls and wire earnings weekly — 0% take-rate for founding sellers.

Learn more →

Pricing

Private beta — join the waitlist.

We're in private beta. Leave your email and we'll reach out when access opens.

Free

$0/mo

For exploration and side projects

100 API calls / day
Live prices endpoint
7-day price history
Compare endpoint
Price alerts
Priority support

What's moving
the inference market.

Curated from 30+ vetted sources. Every story tagged with market signal — so you know what it means for your inference costs.

Loading…

Fetching latest market intelligence…

NEUTRAL

Loading stories…

📰 30-day story archive available on Developer and Team plans.

Infrastructure Pipeline

Global AI capacity
coming online.

Track every major data center announcement, construction phase, and go-live date. New supply = lower prices.

Operator	Project / Location	Region	GPU / Chip	Capacity	CapEx ($B)	Status	Est. Live
Loading infrastructure data…

Data compiled from public announcements and press releases. Submit a correction →

For inference operators

Got spare compute?
Put it to work.

List your inference capacity on the Coaxiom exchange. Set your price. We proxy the calls and wire your earnings weekly.

✓ Any OpenAI-compatible endpoint — set your price
✓ Your credentials never exposed — Coaxiom proxies every request
✓ Weekly payouts via Stripe. 0% take-rate for founding sellers.
✓ Automated uptime monitoring — your listing stays credible

Exchange — Coming Soon

EXAMPLE LISTING
model → meta-llama/llama-3.1-70b
input → $0.18 / 1M
output → $0.35 / 1M
capacity → 500B / month
region → us-east
uptime → 99.4% (30d)
YOUR MONTHLY EARNINGS
$2,430
at 10B served · 0% platform fee

Private beta · Limited access

Get early access to
Coaxiom.

We're in private beta. Leave your email and we'll reach out when your spot opens.

No spam. Unsubscribe any time.

See live prices → Talk to us

The inferenceprice layer.

One call. Every provider.

Real-time fetch

OHLC price history

Price alerts Developer

One clean API

A price API thatactually makes sense.

Got spare inference capacity? Sell it.

Private beta — join the waitlist.

What's movingthe inference market.

Global AI capacitycoming online.

Got spare compute?Put it to work.

Get early access toCoaxiom.