Platform activity
All time0total requests
0tokens routed
$0.0000saved vs official
0accounts
Deposit with USDC on Solana or QRIS. No card on file, no recurring charge.
181 frontier models. The tools you already use.
One key, usage-based pricing, and no monthly commitment.
Set up your agent with AtmoRouter
npx @atmorouter/helperOpenAI-compatible. Pay per token. Explore the docs
181 models 17 families
Explore the catalogue- Avg off official
- 79.8%
- Max context
- 1.1M
Usage across AtmoRouter
Routing health
Measured—success rate, 7d
—median latency
0models routable
0upstream families
From request logs. Failed upstream requests are never billed.
Per-family health appears here as soon as traffic flows.
Explore the model families
Loading capacity…
Routing 181 models across 17 families — up to 100.0% below official pricing
cp/anthropic/claude-opus-4.1From $1.875cc/claude-fable-5From $12.5cc/claude-fable-5-1From $11.25cp/anthropic/claude-fable-5From $12.5cp/openai/gpt-6-astra-proFrom $1.25cx/gpt-6-astraFrom $2.5ag/claude-opus-4-6-thinkingFrom $0.0125cb/claude-opus-4.6From $0.075cb/claude-opus-4.7-1mFrom $0.1cc/claude-opus-4-5From $1.25cc/claude-opus-4-6From $1.25cc/claude-opus-4-7From $1.25cc/claude-opus-4-8From $1.25cc/claude-opus-5From $1.25cp/anthropic/claude-opus-4.5From $0.625cp/anthropic/claude-opus-4.6From $0.625cp/anthropic/claude-opus-4.7From $0.625cp/anthropic/claude-opus-4.8From $4.6875cp/anthropic/claude-opus-4.1From $1.875cc/claude-fable-5From $12.5cc/claude-fable-5-1From $11.25cp/anthropic/claude-fable-5From $12.5cp/openai/gpt-6-astra-proFrom $1.25cx/gpt-6-astraFrom $2.5ag/claude-opus-4-6-thinkingFrom $0.0125cb/claude-opus-4.6From $0.075cb/claude-opus-4.7-1mFrom $0.1cc/claude-opus-4-5From $1.25cc/claude-opus-4-6From $1.25cc/claude-opus-4-7From $1.25cc/claude-opus-4-8From $1.25cc/claude-opus-5From $1.25cp/anthropic/claude-opus-4.5From $0.625cp/anthropic/claude-opus-4.6From $0.625cp/anthropic/claude-opus-4.7From $0.625cp/anthropic/claude-opus-4.8From $4.6875One API.
A lot less in your way.
One key, all models
Drop-in replacement for OpenAI. Change the base URL, keep your SDK. Every model in the catalog answers to a single ar- key.
Your SDKSame request shape
AtmoRouter
atmorouter.dev/v1Authorization: Bearer ar-…One key. Every family.Routed to healthy capacity
Requests land on capacity that is up and within budget. Exhausted or degraded sources are skipped automatically — no manual re-rolls.
RequestHealthy capacity
Exact per-token metering
Input and output priced separately and metered from the upstream usage report, not estimated from characters.
prompt_tokens+completion_tokensStreaming that streams
Full SSE passthrough, identical to OpenAI’s format, with usage on the final frame. Stop the stream, stop paying.
Keys with guardrails
Unlimited keys with per-key rate limits and model allowlists. Revoke one without touching the others.
Usage you can audit
Every request logged with tokens, latency and cost. Filter by model or window, export the raw data anytime.
Keep your code.
Change the endpoint.
Set environment variables, or pass the base URL to your SDK.
Works with most coding agents and CLIs without touching codebash
export OPENAI_BASE_URL="https://atmorouter.dev/v1"
export OPENAI_API_KEY="ar-xxxxxxxxxxxx"
# Any OpenAI-compatible tool now routes through AtmoRouter.
# Verify the key and see every model you can call:
curl $OPENAI_BASE_URL/models -H "Authorization: Bearer $OPENAI_API_KEY"Chat completions, streaming, tool callspython
from openai import OpenAI
client = OpenAI(
base_url="https://atmorouter.dev/v1",
api_key="ar-xxxxxxxxxxxx",
)
res = client.chat.completions.create(
model="cp/anthropic/claude-opus-4.1",
messages=[{"role": "user", "content": "Explain PKCE."}],
)
print(res.choices[0].message.content)Drop-in replacement for the OpenAI clientjavascript
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://atmorouter.dev/v1",
apiKey: process.env.ATMOROUTER_API_KEY,
});
const res = await client.chat.completions.create({
model: "cp/anthropic/claude-opus-4.1",
messages: [{ role: "user", content: "Explain PKCE." }],
});SSE passthrough with exact token usage on the final framepython
stream = client.chat.completions.create(
model="cp/anthropic/claude-opus-4.1",
messages=[{"role": "user", "content": "Count to five"}],
stream=True,
)
for chunk in stream:
delta = chunk.choices[0].delta.content if chunk.choices else None
if delta:
print(delta, end="", flush=True)
# The last frame before [DONE] carries usage:
# {"usage": {"prompt_tokens": 9, "completion_tokens": 25}}
# Billing uses those exact counts, never a character estimate.Base URL
https://atmorouter.dev/v1One host for every model in the catalogue.
Auth header
Authorization: Bearer ar-…x-api-key is accepted too.
Cost header
X-Atmorouter-CostExact amount deducted, on every response.
Find your next model
Prices vary by available offers. Requests can use any listed tier by default. Set maximum input and output prices in Budgets to limit routing.
Flagship model of each familyUSD / million tokens
| Model | Family | Context | Input $/Mtok | Output $/Mtok | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
cp/anthropic/claude-opus-4.15 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | Cline Pass | 200K | From $1.875 | From $9.375 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
cc/claude-fable-51 price tierFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | Claude Code | 1M | From $12.5 | From $62.5 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
cx/gpt-6-astra9 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | Codex | 272K | From $2.5 | From $12.5 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ag/claude-opus-4-6-thinking28 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | Antigravity | 1M | From $0.0125 | From $0.0625 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
cb/claude-opus-4.614 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | CodeBuddy | 1M | From $0.075 | From $0.375 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
kiro/claude-opus-4.56 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | KIRO | 200K | From $0.6125 | From $3.0625 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
grok/grok-4.7-build-fast9 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | GROK | 500K | From $0.08 | From $0.24 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
kcai/k3.1-pro-highspeed2 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | KCAI | 262K | From $0.99 | From $4.95 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
mai/muse-wonder-1.05 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | MAI | 1.0M | From $0.1 | From $0.5 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ali/kimi-k38 price tiersFinal prices in USD per million tokens. Each tier is compared with official input and output prices.
Offer counts are a snapshot and do not guarantee available capacity. Updated | Alibaba Cloud | 1M | From $0.0075 | From $0.0375 |
Simple, usage-based pricing
Pay as you go
Per million tokens, in USD
- One key for every model
- Input and output priced separately
- Real-time balance deduction
- Stop calling, stop paying
Built for agents
Long-running, high-volume workloads
- Full SSE streaming passthrough
- Per-key rate limits and allowlists
- Prompt caching where supported
- Latency and token logs per request
Up to 100.0% off
Versus official list pricing
- Frontier models without the plan
- Transparent per-model rates
- Failed requests are never billed
- Prices refresh continuously
What can $1 get you? Explore the Token Calculator →
From account to first request
Create account
Sign up with an email. Nothing else required to look around.
Add balance
Deposit USDC on Solana or pay with QRIS. Balance drains per token, never per month.
Generate a key
Create an ar- key, set its rate limit and allowed models.
Start calling
Point any OpenAI-compatible client at atmorouter.dev/v1.
How your usage is billed
Metering comes from the upstream usage report, so the number you are charged is the number you consumed.
What you get
- One key for the OpenAI-compatible endpoint
- 181 models across 17 families
- Real-time balance deduction per request
- Full SSE streaming with usage on the final frame
- Per-key rate limits and model restrictions
- Usage dashboard with cost and latency breakdown
How billing works
- 01
Deposit USDC on Solana, or top up with QRIS
- 02
Input and output tokens counted separately
- 03
Cached input bills at 10% of the input rate
- 04
cost = ((in − cached) × input_price + cached × input_price × 0.10 + out × output_price) / 1M
- 05
Balance deducted the moment a request settles
- 06
A failed upstream request is never billed
- 07
Empty balance blocks new calls, never overdraws
Create an account, grab a key, and call 181 models in minutes.
Community
A direct line
to AtmoRouter.
Get integration help, request models, and follow service updates.
- Integration help from us directly
- Model requests and pricing feedback
- Incident and maintenance notices
- Early access to new features