Token prices are hard to compare across providers — enter your usage once and see the real monthly bill for every major model.
Live data · updated 2026-10-30 10:03 · sources: LMArena + OpenRouter
For a typical indie workload (50M in / 10M out per month), the spread between the cheapest and most expensive frontier model is over 40× — most apps can run on Gemini 3.8 Flash (High) for single-digit dollars.
| # | Model | Arena | $/1M in | $/1M out | Monthly cost |
|---|---|---|---|---|---|
| 1 | Gemini 3.8 Flash (High)Google | 1496 | $0.75 | $3.75 | $75.00 |
| 2 | Muse Spark 1.3 (Max)Meta | 1495 | $1.25 | $4.25 | $105.00 |
| 3 | Muse Spark 1.2 (xHigh)Meta | 1492 | $1.25 | $4.25 | $105.00 |
| 4 | Gemini 4 Carbon (High)Google | 1528 | $2 | $10 | $200.00 |
| 5 | Gemini 4 Argon (High)Google | 1520 | $2 | $10 | $200.00 |
| 6 | Claude Opus 5.5 (High)Anthropic | 1509 | $4 | $20 | $400.00 |
| 7 | Muse Wonder (Max)Meta | 1507 | $4 | $20 | $400.00 |
| 8 | Claude Opus 4.6 (High)Anthropic | 1503 | $5 | $25 | $500.00 |
| 9 | Claude Opus 4.7 (High)Anthropic | 1501 | $5 | $25 | $500.00 |
| 10 | Claude Opus 4.6Anthropic | 1496 | $5 | $25 | $500.00 |
| 11 | Claude Opus 4.7Anthropic | 1494 | $5 | $25 | $500.00 |
| 12 | Claude Opus 5 (High)Anthropic | 1492 | $5 | $25 | $500.00 |
| 13 | Claude Opus 5 (Max)Anthropic | 1491 | $5 | $25 | $500.00 |
| 14 | Claude Fable 5.1 (Max)Anthropic | 1504 | $10 | $50 | $1,000 |
| 15 | Claude Fable 5 (High)Anthropic | 1504 | $10 | $50 | $1,000 |
50M input + 10M output tokens per month. Cheapest: Gemini 3.8 Flash (High) at $75.00. Prices from OpenRouter, arena ratings from LMArena.
One English word ≈ 1.3 tokens; a typical chat turn (your prompt + history) runs 500–2,000 input tokens. RAG pipelines multiply input fast — every retrieval re-sends context. Measure one week of real traffic before committing to a model tier.
As of 2026-10-30 10:03, output tokens cost 3–5× more than input on most models. Chat-heavy products should weight output price; summarization and RAG products should weight input. Batch APIs (off-peak) cut costs ~50% on several providers if latency is flexible.
A model 20% cheaper that fails 5% more often costs more once you count retries and bad user experiences. Use the arena rating column as a quality floor, then optimize price above it.