Deployment-level pricing, capabilities, lifecycle and provider compatibility—normalized, sourced and kept current.
Public registry
The same model is not always the same product across providers.
| Model | Deployment | Input price | Output price | Context | Capability | Lifecycle |
|---|---|---|---|---|---|---|
| Loading the live catalog… | ||||||
Live intelligence reports
Which Azure AI model versions actually retire in October 2026 — and where the live registry currently disagrees with Microsoft's schedule.
Open reportA live watchlist of the provider endpoints ALLM has observed in the last seven days.
Open reportSee concrete deployment retirement dates before they become a migration incident.
Open reportProduction workflows
Resolve, compare, and price provider deployments with one normalized intelligence API.
Explore the implementation guidesModel EOL
See announced EOL dates, date confidence, and replacements early enough to plan the migration.
Check upcoming EOL datesDeployment
Claude Haiku 4.5
Anthropic · Messages · Global
End of life
Agent reliability
Precompute candidates that preserve the capabilities your agent cannot lose before your gateway needs them.
Build a failover planAnthropic · Messages
dep_anthropic_claude_haiku_4_5_messages_global
AWS Bedrock · Converse
dep_aws_bedrock_claude_haiku_4_5_converse_eu
Gateway route · Chat
dep_gateway_claude_haiku_4_5_chat_eu
AI SaaS economics
Quote one product action from its exact deployment and expected input, cached-input, and output tokens.
Price a product actionDeployment
GPT-5.1 · Responses
Total cost
Margin rules stay in your product.
$2.250000
Routing policy
Reject incompatible capabilities first, then compare cost only across the deployments that qualify.
See constraint routingcurl https://api.use-allm.com/v1/deployments/compare \
-X POST -d '{
"deployment_ids": ["dep_anthropic_…", "dep_bedrock_…"],
"usage": { "input_tokens": 1000000, "output_tokens": 100000 },
"requires": {
"tool_calling": "supported",
"structured_output": "conditional"
}
}'Qualified
2 deployments
Ranking
public token cost
FinOps & billing
Join your usage events to canonical deployment identities and normalized public-rate quotes.
Build a cost ledgerYour event
customer_id
cus_acme
feature
support_agent
input_tokens
1000000
output_tokens
100000
+ ALLM lookup
deployment_id
dep_openai_gpt_5_1_…
estimated_cost_usd
2.250000
catalog_release
cat_2026_07_…
Model identity
Resolve a canonical model into the exact provider deployment your logs, billing, and incident tools can record.
Resolve model identityCanonical release
anthropic/claude-haiku-4.5
Anthropic
claude-haiku-4-5
Messages · Global
Direct
AWS Bedrock
anthropic.claude-haiku-4-5-v1:0
Converse · EU
Cloud
ALLM provides deployment intelligence. Your stack sends requests, records usage, and executes decisions.
Cached tokens, regional uplifts, batch discounts and private price books are applied with an auditable quote.
const quote = await allm.pricing.calculate({
deployment: "dep_openai_gpt_5_1_responses_global",
inputTokens: 1_000_000,
outputTokens: 100_000,
});
console.log(quote.total); // "$2.25"Beyond a registry
The public data stays accessible. Teams pay for scale and control.
We use technologies necessary for the website to operate. With your consent, we use PostHog and Google Analytics to measure pseudonymised audience data. Refusal does not affect access to the website.