TTokenCost

Compare LLM pricing

The model matchups people search for most, priced on the same workload: 1,000 input tokens and 500 output tokens per request.

GPT-5 vs Claude Sonnet 4.5

GPT-5cheapest$0.0063
Claude Sonnet 4.5$0.01

GPT-5 is 1.7x cheaper on this workload. Context differs: 400K vs 200K.

GPT-5 mini vs Gemini 2.5 Flash

GPT-5 minicheapest$0.0013
Gemini 2.5 Flash$0.0015

GPT-5 mini is 1.2x cheaper on this workload. Context differs: 400K vs 1049K.

GPT-4o vs Claude Haiku 4.5

GPT-4o$0.0075
Claude Haiku 4.5cheapest$0.0035

Claude Haiku 4.5 is 2.1x cheaper on this workload. Context differs: 128K vs 200K.

DeepSeek V3.2 (chat) vs GPT-4o mini

DeepSeek V3.2 (chat)$0.0005
GPT-4o minicheapest$0.0004

GPT-4o mini is 1.1x cheaper on this workload.

DeepSeek R1 (reasoner) vs GPT-5

DeepSeek R1 (reasoner)cheapest$0.0016
GPT-5$0.0063

DeepSeek R1 (reasoner) is 3.8x cheaper on this workload. Context differs: 128K vs 400K.

Gemini 2.5 Pro vs GPT-5

Gemini 2.5 Procheapest$0.0063
GPT-5$0.0063

Gemini 2.5 Pro is 1.0x cheaper on this workload. Context differs: 1049K vs 400K.

Llama 4 Maverick vs DeepSeek V3.2 (chat)

Llama 4 Maverick$0.0007
DeepSeek V3.2 (chat)cheapest$0.0005

DeepSeek V3.2 (chat) is 1.4x cheaper on this workload. Context differs: 1049K vs 128K.

Claude Opus 4.1 vs Claude Sonnet 4.5

Claude Opus 4.1$0.05
Claude Sonnet 4.5cheapest$0.01

Claude Sonnet 4.5 is 5.0x cheaper on this workload.

Want a pairing that is not listed?

The full pricing table includes every model we track, and the calculator lets you pick any two models and compare against your real prompt length.

Full pricing tableOpen calculator