LLM API pricing, verified weekly
Read the market
like paper.
Registry 13 models · Refresh weekly · Sponsored none
Every major LLM API — input, output and cached-token prices per 1M tokens, context windows and capabilities. Verified against official pages, refreshed weekly, free forever.
Cheapest right now
updated Aug 19, 2026| # | Model | Input $/1M | Output $/1M |
|---|---|---|---|
| 01 | DeepSeek V4 Pro Open | $0.660† | $1.98 |
| 02 | Muse Spark 1.2 Closed | $1.25 | $4.25 |
| 03 | GLM-5.3 Closed | $1.40 | $4.40 |
| 04 | Grok 4.6 Closed | $2.00 | $6.00 |
| 05 | Qwen3.8-Max Closed | $2.00 | $6.00 |
† off-peak rate; peak windows bill 2×
Full table →Models tracked
13
▲ open + closed weights
Min input $/1M
$0.660
▼ cheapest verified rate
Max context
2M
▲ tokens in one window
Providers
10
▲ official sources only
Tools
T/01
Pricing
Full table of model × provider prices per 1M tokens. Sort, filter by capability, check freshness.
T/02
Calculator
Plug in your traffic — requests per day and token volumes — and get the monthly bill per model.
T/03
Guides
Practical guides: how to compare API pricing, cut costs with caching, pick a model for a task.
T/04
Benchmarks
LMArena Elo ratings next to SWE-bench, GPQA, AIME and more — every score sourced, caveats flagged.