Compare provider prices for the same language model.
Compare the cost of calling the same language model through different providers. This view joins complete input and output quotes under one exact model identity, so you can inspect a stated workload without collecting each rate card yourself.
217 exact language-model identities have complete recorded token quotes from at least two providers. Up to four models and 20 routes are shown. The example workload uses 1M input and 250,000 output tokens, without cache or batch discounts.
Catalog snapshot · 11 rows shownSet the input and output quantities
The example uses one million input tokens and 250,000 output tokens. Its charge is the input rate plus one quarter of the output rate, using both legs from the same route. It excludes prompt caching, batch discounts, retries and other charges not represented by those two rates.
Your workload can have a different balance. An input-heavy document task and a long generated answer can favor different rate structures. Keep both prices in view rather than comparing only the lower input number. An overdue or ineligible route has no current workload estimate here.
Check that the serving conditions fit
The catalog identity joins explicitly resolved model variants. That still leaves endpoint context limits, configuration, region, availability and provider terms to check. The table retains reported configuration and endpoint context instead of assuming identical service from a matching model name.
The rows are not an overall provider ranking. No latency, reliability or privacy result is inferred from a token price. Open the price source and model routes, then test the provider configuration you would actually use. Missing output rates cannot produce a complete token-workload estimate.
How this view is assembled
Selection: exact language models with complete non-negative input/output token quotes from at least two provider IDs. Conditional, degraded, component and superseded aggregate quotes are excluded. Up to four models and five routes each. Models with current quotes appear first. Each table takes one route per provider before additional configurations, favoring current eligibility and stable provider order. Current estimates require the explorer’s freshness and future-date checks.
Amazon: Nova 2 Lite
One exact catalog model; provider settings and access terms can still differ. Current quotes appear first, followed by stable provider order; this is not a winner ranking.
| Model | Provider rate | Example workload | Configuration and status | Sources and dates |
|---|---|---|---|---|
| Amazon: Nova 2 Lite | Amazon Bedrock$0.30 input / $2.50 output per 1M tokensUSD | $0.9251M input + 250K output tokens | amazon-bedrock1M endpoint contextCurrent eligible quote | |
| Amazon: Nova 2 Lite | Amazon Bedrock$0.30 input / $2.50 output per 1M tokensUSD | $0.9251M input + 250K output tokens | amazon-bedrock/eu-west-11M endpoint contextCurrent eligible quote |
Amazon: Nova Lite 1.0
One exact catalog model; provider settings and access terms can still differ. Current quotes appear first, followed by stable provider order; this is not a winner ranking.
| Model | Provider rate | Example workload | Configuration and status | Sources and dates |
|---|---|---|---|---|
| Amazon: Nova Lite 1.0 | Amazon Bedrock$0.06 input / $0.24 output per 1M tokensUSD | $0.121M input + 250K output tokens | amazon-bedrock300K endpoint contextCurrent eligible quote | |
| Amazon: Nova Lite 1.0 | Amazon Bedrock$0.06 input / $0.24 output per 1M tokensUSD | $0.121M input + 250K output tokens | amazon-bedrock/eu-west-1300K endpoint contextCurrent eligible quote |
Amazon: Nova Pro 1.0
One exact catalog model; provider settings and access terms can still differ. Current quotes appear first, followed by stable provider order; this is not a winner ranking.
| Model | Provider rate | Example workload | Configuration and status | Sources and dates |
|---|---|---|---|---|
| Amazon: Nova Pro 1.0 | Amazon Bedrock$0.80 input / $3.20 output per 1M tokensUSD | $1.601M input + 250K output tokens | amazon-bedrock300K endpoint contextCurrent eligible quote | |
| Amazon: Nova Pro 1.0 | Amazon Bedrock$0.80 input / $3.20 output per 1M tokensUSD | $1.601M input + 250K output tokens | amazon-bedrock/eu-west-1300K endpoint contextCurrent eligible quote |
Anthropic: Claude Fable 5.1
One exact catalog model; provider settings and access terms can still differ. Current quotes appear first, followed by stable provider order; this is not a winner ranking.
| Model | Provider rate | Example workload | Configuration and status | Sources and dates |
|---|---|---|---|---|
| Anthropic: Claude Fable 5.1 | Amazon Bedrock$10.00 input / $50.00 output per 1M tokensUSD | $22.501M input + 250K output tokens | amazon-bedrockCache write $12.5/M tokens · 1h cache write $20/M tokens · Web search $0.01/search1M endpoint contextCurrent eligible quote | |
| Anthropic: Claude Fable 5.1 | Anthropic$10.00 input / $50.00 output per 1M tokensUSD | $22.501M input + 250K output tokens | anthropicCache write $12.5/M tokens · 1h cache write $20/M tokens · Web search $0.01/search1M endpoint contextCurrent eligible quote | Anthropic price source ↗Checked |
| Anthropic: Claude Fable 5.1 | Azure$10.00 input / $50.00 output per 1M tokensUSD | $22.501M input + 250K output tokens | azureCache write $12.5/M tokens · 1h cache write $20/M tokens · Web search $0.01/search1M endpoint contextCurrent eligible quote | Azure price source ↗Checked |
| Anthropic: Claude Fable 5.1 | Experiential$10.00 input / $50.00 output per 1M tokensUSD | $22.501M input + 250K output tokens | Standard list rateCatalog list rate. Plan allowances and temporary discounts are separate.1M endpoint contextCurrent eligible quote | Experiential price source ↗Checked |
| Anthropic: Claude Fable 5.1 | Google$10.00 input / $50.00 output per 1M tokensUSD | $22.501M input + 250K output tokens | google-vertex/globalCache write $12.5/M tokens · 1h cache write $20/M tokens · Web search $0.01/search1M endpoint contextCurrent eligible quote | Google price source ↗Checked |
Where these rows come from
We join catalog specifications, recorded provider quotes and published evaluation records where exact model identities are available. Some observations come through aggregators or benchmark collections. Multiple listings do not establish independent confirmation of a claim.
Open the source beside a row to check the original evidence. A collection or check date is separate from a model release or evaluation date. Unknown values stay unknown. Recorded prices are in USD with their original billing units and settings.
See data coverage and sources →Compare for your own use case
Keep your task, model identity and provider configuration together. Explore the catalog yourself, or connect your AI to read the same source-backed records.