All plans · Moonshot AI
Kimi Max: $99/month
Kimi Max covers about 3.3 turns of balanced coding work a month, 0.034 per dollar. That is the 24th most of the 37 plans with a usage figure. At $99 or less, Claude Pro covers the most: about 1.1k.
What it includes
| Price | $99/month |
|---|---|
| Coding agent | Kimi Code |
| Limit, as Moonshot AI words it | Rolling five-hour limit and a monthly pool. |
| Usage at API prices | $720–$1.3k a month, 7.3–13× the price |
| Evidence | Measured · checked Nov 12, 2026 |
| Strongest model | Kimi K3.1 Pro, 46.5 on the Intelligence Index |
Models
- Kimi K3.1 Pro · up to 46.5
- Kimi K3 · up to 17.4 · About 10× K3.1 Pro's tokens per window
Value by kind of work
Turns of mixed coding work the allowance covers in a month, each kind on the plan’s cheapest model that is good enough. How this is worked out.
| Work mix | Turns a month | Per dollar |
|---|---|---|
| Light only (100% light) | 61–109 | 0.62–1.1 |
| Mostly light (40% normal, 60% light) | 61–109 | 0.62–1.1 |
| Balanced (5% frontier, 15% hard, 60% normal, 20% light) | 2.5–4.4 | 0.025–0.045 |
| Mostly hard (10% frontier, 40% hard, 40% normal, 10% light) | 1.2–2.1 | 0.012–0.021 |
On the balanced mix: frontier work on Kimi K3.1 Pro (default reasoning), counting 0% of a turn; hard work on Kimi K3.1 Pro (default reasoning), counting 3% of a turn; normal work on Kimi K3.1 Pro (default reasoning); light work on Kimi K3.1 Pro (default reasoning).
How the figure was obtained
Moonshot publishes no numbers. From one unofficial measurement (wokey.ai, November 2026, a reseller of these plans): tokens per five-hour window on Kimi K3.1 Pro, about 12 windows a month, valued at API prices with 93–99% cache hits.
Compare
Other Kimi tiers: Kimi Plus ($19) · Kimi Pro ($39) · Kimi Ultra ($199)
At a similar price: Claude Team Premium ($100, about 8k turns) · Claude Max ($100, about 6.9k turns) · Google AI Ultra 5× ($99.99, about 2.8k turns) · ChatGPT Pro 100 ($100, about 2.1k turns)
See every plan on the charts, or the best plan at each budget.