What is the cheapest LLM API?
It depends on your input/output token mix. Sort by blended price here, then confirm the model has the context window and capabilities your workload needs.
/ tools / australia / llm-price-compare
LLM Price Compare
Compare model price, context, source, and capabilities before routing traffic.
Compare LLM API prices side by side — input and output cost per million tokens, blended price, context window, capabilities, and data source — before you route production traffic. Includes domestic and global providers so you can weigh region fit alongside cost.
- Rank models by blended price per million tokens
- Check context window, tool/JSON/vision support, and production fit
- Spot the cheapest model that still meets your capability needs
Price Compare
Inputs are processed in this browser; do not paste real API keys.
Diagnostic report
- rows
- [{"slug":"anthropic--claude-fable-5","name":"Anthropic: Claude Fable 5","providerName":"Anthropic","currency":"USD","inputPricePerM":1.4467592592592593,"outputPricePerM":7.233796296296297,"blendedPerM":8.68055556,"contextLength":1000000,"supportsTools":true,"supportsJson":true,"supportsVision":true,"supportsStreaming":true,"productionFit":"catalog_review","regionFit":"global_only","sourceLabel":"NextModel gateway catalog (Go origin)"},{"slug":"anthropic--claude-haiku-4.5","name":"Anthropic: Claude Haiku 4.5","providerName":"Anthropic","currency":"USD","inputPricePerM":0.14467592592592593,"outputPricePerM":0.7233796296296297,"blendedPerM":0.86805556,"contextLength":200000,"supportsTools":true,"supportsJson":true,"supportsVision":true,"supportsStreaming":true,"productionFit":"catalog_review","regionFit":"global_only","sourceLabel":"NextModel gateway catalog (Go origin)"}]
- cheapestSlug
- anthropic--claude-haiku-4.5
- warnings
- ["some_models_need_catalog_review_before_production"]
some_models_need_catalog_review_before_production
FAQ
LLM Price Compare FAQ
How should I compare model prices fairly?
Use a blended cost that weights input and output by your real ratio, and normalize currencies. Cheap output but expensive input can flip the ranking for chat-heavy traffic.
Are these prices live?
Prices come from the NextModel catalog with a source label per model. Always confirm against the provider before committing high-volume traffic.
Related tools
Keep optimizing cost
Next step
Use the report as the next integration decision.
Copy the base URL, compare model details, or create a key when you are ready to run a real compatibility test.