22 Models · OpenAI, Anthropic, Google, DeepSeek, xAI, Qwen & Mistral
AI API Token & Cloud Cost Simulator
Pick a workload preset or paste your prompt, dial in prompt caching hit rates, and compare real monthly bills across 22 leading AI models.
Quick Workload Presets
10,000 / day (300,000/mo)
2,500 tokens
600 tokens
50% cached
Automatically updates input tokens above using real-time length estimation.
Monthly Cost Comparison (22 Models · Cheapest First)
| Model | Tier | Input / Output (1M) | Monthly Bill | Cache Savings | Deploy / Credits |
|---|---|---|---|---|---|
| GPT-6 LunaOpenAI | Budget | $0.10 / $0.50 | $131.25 | -$33.75 | Deploy via OpenRouter |
| Claude Haiku 5.5Anthropic | Budget | $0.10 / $0.50 | $131.25 | -$33.75 | Claim Cloud Credits |
| DeepSeek V4.1 FlashDeepSeek | Budget | $0.15 / $0.60 | $165.38 | -$55.13 | Route via OpenRouter |
| GPT-OSS 120B (Groq)Groq | Budget | $0.15 / $0.60 | $178.13 | -$42.38 | Run Fast on Groq |
| Qwen 3.7 PlusQwen | Workhorse | $0.32 / $1.28 | $362.40 | -$108.00 | Route via OpenRouter |
| Gemini 3.1 Flash-LiteGoogle | Budget | $0.25 / $1.50 | $373.13 | -$84.38 | Claim $300 GCP Credits |
| Qwen 3.6 FlashQwen | Budget | $0.25 / $1.50 | $373.13 | -$84.38 | Route via OpenRouter |
| Mistral Large 3Mistral | Workhorse | $0.50 / $1.50 | $476.25 | -$168.75 | Route via OpenRouter |
| Gemini 3.5 Flash-LiteGoogle | Budget | $0.30 / $2.50 | $573.75 | -$101.25 | Claim $300 GCP Credits |
| DeepSeek V4 ProDeepSeek | Workhorse | $0.66 / $1.98 | $612.15 | -$239.25 | Route via OpenRouter |
| Grok 4.20xAI | Workhorse | $1.25 / $2.50 | $965.63 | -$421.88 | Route via OpenRouter |
| Gemini 3.8 FlashGoogle | Workhorse | $0.75 / $3.75 | $984.38 | -$253.13 | Claim $300 GCP Credits |
| GPT-6.1 TerraOpenAI | Workhorse | $1.00 / $5.00 | $1,312.50 | -$337.50 | Route via OpenRouter |
| Mistral Large 4Mistral | Frontier | $1.36 / $4.18 | $1,313.40 | -$459.00 | Route via OpenRouter |
| Grok 4.5xAI | Frontier | $2.00 / $6.00 | $1,905.00 | -$675.00 | Route via OpenRouter |
| Qwen 3.7 MaxQwen | Frontier | $2.50 / $7.50 | $2,381.25 | -$843.75 | Route via OpenRouter |
| GPT-6.1 SolOpenAI | Workhorse | $2.00 / $10.00 | $2,587.50 | -$712.50 | Route via OpenRouter |
| Claude Sonnet 5.5Anthropic | Workhorse | $2.00 / $10.00 | $2,587.50 | -$712.50 | Deploy on Cloud |
| Gemini 4 ArgonGoogle | Frontier | $2.00 / $10.00 | $2,587.50 | -$712.50 | Claim $300 GCP Credits |
| Claude Opus 5.5Anthropic | Frontier | $4.00 / $20.00 | $5,175.00 | -$1425.00 | Deploy on Cloud |
| GPT-6 AstraOpenAI | Frontier | $10.00 / $50.00 | $13,125.00 | -$3375.00 | Route via OpenRouter |
| Claude Fable 5.1Anthropic | Frontier | $10.00 / $50.00 | $13,125.00 | -$3375.00 | Deploy on Cloud |
AI API Pricing & Prompt Caching FAQ
How are LLM API tokens calculated?
For English text and code, 1,000 tokens is roughly 750 words or 4,000 characters. Providers bill input tokens (your prompt and context) and output tokens (the generated reply) separately per 1 million tokens.
How much does Prompt Caching save on AI API bills?
Prompt caching discounts repeated system instructions, tools, or RAG context by up to 90% on input tokens across OpenAI, Anthropic, and Google models.
When should I use the Async Batch API?
If your workload does not require instant real-time responses (such as overnight data extraction, classification, or bulk summarization), Batch APIs cut token costs by 50%.
Popular Model Cost Comparisons (231 Total Pairs)
Browse all 231 comparisons →GPT-6 Luna vs DeepSeek V4.1 Flash →GPT-6.1 Sol vs Claude Sonnet 5.5 →Claude Sonnet 5.5 vs Gemini 4 Argon →GPT-6.1 Sol vs Grok 4.5 →DeepSeek V4.1 Flash vs Qwen 3.6 Flash →DeepSeek V4 Pro vs Mistral Large 3 →GPT-6 Astra vs Claude Opus 5.5 →GPT-6.1 Terra vs Gemini 3.8 Flash →Claude Haiku 5.5 vs DeepSeek V4.1 Flash →Grok 4.20 vs Mistral Large 4 →Gemini 3.8 Flash vs Qwen 3.7 Plus →GPT-6 Luna vs GPT-OSS 120B (Groq) →Claude Opus 5.5 vs Gemini 4 Argon →Claude Sonnet 5.5 vs Qwen 3.7 Max →GPT-6.1 Sol vs DeepSeek V4 Pro →