LLM API cost calculator
Enter your monthly request volume and average token counts — we rank every major LLM API by what it would actually cost you per month.
| Model | $/1M in | $/1M out | Context | Monthly | Notes |
|---|---|---|---|---|---|
| Claude Haiku 5.5 Anthropic | $0.1 | $0.5 | 1000K | $45.00 | Released 2026-10-07; prompts >100K tokens $0.50/$2.50 (whole request); cache read $0.01 |
| GPT-6 Luna OpenAI | $0.1 | $0.5 | 1050K | $45.00 | Cached input $0.01 |
| Mistral Small 4 Mistral | $0.15 | $0.6 | 256K | $60.00 | |
| GPT-5.6 Luna OpenAI | $0.2 | $1.2 | 1050K | $100 | |
| DeepSeek V4.1 Flash DeepSeek | $0.3 | $1.2 | 1000K | $120 | API id deepseek-flash; off-peak −50%; cache-hit input $0.006 |
| Mistral Large 3 Mistral | $0.5 | $1.5 | 256K | $175 | Open-weight |
| Gemini 3.8 Flash Google | $0.75 | $3.75 | 1000K | $338 | Promo price through 2026; $1.50/$7.50 from 2027 |
| Gemini 3.7 Flash Google | $0.75 | $3.75 | 1000K | $338 | Deprecated 2026-10-08, requests auto-route to Gemini 3.8 Flash; no longer on the pricing page |
| GPT-5.4 mini OpenAI | $0.75 | $4.5 | 400K | $375 | Legacy (not in current model list) |
| Grok 4.3 xAI | $1.25 | $2.5 | 1000K | $375 | ≤200k prompt; higher above |
| Claude Haiku 4.5 Anthropic | $1 | $5 | 200K | $450 | Previous generation (superseded by Haiku 5.5) |
| DeepSeek V4 Pro DeepSeek | $1.32 | $3.96 | 1000K | $462 | Reasoning tier; off-peak −50% |
| Mistral Medium 3.5 Mistral | $1.5 | $7.5 | 256K | $675 | |
| Grok 4.7 xAI | $2 | $6 | 500K | $700 | Current flagship; ≤200k prompt; higher above |
| Grok 4.6 xAI | $2 | $6 | 500K | $700 | ≤200k prompt; higher above |
| Gemini 3.5 Flash Google | $1.5 | $9 | 1000K | $750 | Deprecated 2026-10-08, requests auto-route to Gemini 3.6/3.8 Flash; no longer on the pricing page |
| Claude Sonnet 5.5 Anthropic | $2 | $10 | 1000K | $900 | Released 2026-09-28; 128K max output |
| Claude Sonnet 5 Anthropic | $2 | $10 | 1000K | $900 | Previous generation (superseded by Sonnet 5.5, same price) |
| GPT-6.1 Sol OpenAI | $2 | $10 | 1050K | $900 | Released 2026-09-30; cached input $0.10; >272k prompt $4/$15 |
| GPT-6 Sol OpenAI | $2 | $10 | 1050K | $900 | Superseded by GPT-6.1 Sol (same price); cached input $0.20; >272k prompt: 2× input, 1.5× output |
| GPT-5.6 Terra OpenAI | $2 | $12 | 1050K | $1,000 | |
| Gemini 3.1 Pro Google | $2 | $12 | 1000K | $1,000 | Preview; ≤200k prompt ($4/$18 above) |
| Claude Opus 5.5 Anthropic | $4 | $20 | 1000K | $1,800 | Current Opus; cache read $0.20 |
| GPT-5.6 Sol OpenAI | $4 | $20 | 1050K | $1,800 | Promo price at least through 2026-11-21 (was $5/$30); cached input $0.40; >272k prompt: 2× input, 1.5× output |
| Claude Opus 5 Anthropic | $5 | $25 | 1000K | $2,250 | Previous generation (superseded by Opus 5.5) |
| GPT-5.5 OpenAI | $5 | $30 | 1050K | $2,500 | Cached input $0.50; legacy (not in current model list) |
| Claude Fable 5.1 Anthropic | $10 | $50 | 1000K | $4,500 | Most capable; cache read $0.25 |
| Claude Fable 5 Anthropic | $10 | $50 | 1000K | $4,500 | Previous generation (superseded by Fable 5.1, same price) |
| GPT-6 Astra OpenAI | $10 | $50 | 1050K | $4,500 | Cached input $1; >272k prompt: 2× input, 1.5× output |
List prices, no caching/batch discounts applied. Data as of 2026-10-12.