Skip to main content

Model identifiers

Use these exact strings as the model parameter in your API requests.

Tier 1: Flagship

ModelProviderInput / 1MOutput / 1MBest for
gpt-4oOpenAI$2.50$10.00General purpose
claude-sonnet-4-6Anthropic$3.00$15.00Coding, analysis
gemini-2.5-proGoogle$1.25$10.00Long context
gpt-4.1OpenAI$2.00$8.00Coding, instructions

Tier 2: Cost-efficient

ModelProviderInput / 1MOutput / 1MBest for
gpt-4o-miniOpenAI$0.15$0.60High volume
gpt-4.1-miniOpenAI$0.40$1.60Balanced
claude-haiku-4-5-20251001Anthropic$0.80$4.00Fast tasks
gemini-2.5-flashGoogle$0.15$0.60Speed

Tier 3: Budget

ModelProviderInput / 1MOutput / 1MBest for
gpt-4.1-nanoOpenAI$0.10$0.40Classification
meta-llama/llama-4-scoutMeta$0.15$0.40Open source
meta-llama/llama-4-maverickMeta$0.20$0.60Open source

Tier 4: Reasoning

ModelProviderInput / 1MOutput / 1MBest for
claude-opus-4-6Anthropic$15.00$75.00Complex reasoning
gpt-4-turboOpenAI$10.00$30.00Legacy
deepseek/deepseek-r1DeepSeek$0.50$2.00Chain-of-thought
deepseek/deepseek-chat-v3-0324DeepSeek$0.50$1.50General

Unknown models

If you use a model not in our pricing table, we apply a default rate of 1.00/1Minputand1.00/1M input and 3.00/1M output for savings calculations. The request still works — it’s routed through OpenRouter normally.