AI Model Pricing Calculator
Compare real costs across 43 models. Thinking tokens, prompt caching, batch pricing — all factored in. 🆕 Claude Opus 5.5, GPT-6 Sol/Luna, and Grok 4.7 added. Updated September 25, 2026.
90%
0.50×
#
Model
Provider
Input $
Output $
Total
Cost Bar
vs GPT-5.5
Pricing sources: DeepSeek · OpenAI · Anthropic · Google · xAI · Kimi K3 · Meta Model API · OpenRouter (verified September 25, 2026; standard short-context input/output rates). 🆕 Kimi K3: $3 input, $15 output, $0.30 cached input per 1M tokens. 🆕 Muse Spark 1.1: $1.25 input, $4.25 output, $0.15 cached input. Claude Sonnet 5: Introductory $2/$10 per 1M through Aug 31, 2026, then $3/$15. DeepSeek: permanent 75% discount. Qwen 3.7 Max: 50% promo thru June 22. GPT-5.6 (standard short context): Sol $4/$20, Terra $2/$12, Luna $0.20/$1.20 per 1M; cached input is 90% discounted. Sol’s $4/$20 rate is promotional. New models: Opus 5.5 $4/$20 (cached $0.20), GPT-6 Sol $2/$10 (cached $0.20), GPT-6 Luna $0.10/$0.50 (cached $0.01), Grok 4.7 $2/$6 (cached $0.50) per 1M input/output tokens. Long-context surcharges are not modeled. Thinking tokens billed as output.
Test these models on your own code
43 models compared live. Side-by-side testing. Find the best model for your workflow.
🚀 Try on CodingFleet →