Kimi K3 cut across 2 hosts, by up to 23% at InferenceNet (input)
Published 26 September 2026
Price move, per 1M tokens
- InferenceNet · input↓ 23%$1.30 → $1.00
- InferenceNet · output↓ 21%$11.40 → $9.00
- InferenceNet · cache readunchanged$0.30 → $0.30
- Sail Research · input↓ 14%$1.20 → $1.03
- Sail Research · output↓ 14%$10.53 → $9.04
- Sail Research · cache readunchanged$0.30 → $0.30
| host | rate | was | now | change |
|---|---|---|---|---|
| InferenceNet | input | $1.30 | $1.00 | ↓ 23% |
| InferenceNet | output | $11.40 | $9.00 | ↓ 21% |
| InferenceNet | cache read | $0.30 | $0.30 | — |
| Sail Research | input | $1.20 | $1.03 | ↓ 14% |
| Sail Research | output | $10.53 | $9.04 | ↓ 14% |
| Sail Research | cache read | $0.30 | $0.30 | — |
InferenceNet input: ↓ 23%Sail Research input and output: ↓ 14%
The model
Kimi K3open the model →
Open weights2779.9B parameters1049K context25 hosts
intelligence12th of 168 writing18th of 168 coding6th of 168 agents10th of 55
Kimi K3 is Moonshot AI's massive-scale mixture-of-experts model with 2.8 trillion total parameters and 104 billion active per word. It excels at reasoning and web-development tasks, and handles up to one million tokens in a single request, though its weights carry commercial restrictions.
Source
Read on openrouter.ai · we published this on 26 September 2026.
read the original at openrouter.ai ↗