Compare
Kimi K3 vs Qwen 3.8 Max: price comparison
Kimi K3 lists at $3 / $15 and Qwen 3.8 Max at $1.65 / $4.95 per 1M input/output tokens. Qwen 3.8 Max is cheaper on both input and output tokens.
List prices
| Kimi K3 | Qwen 3.8 Max | |
|---|---|---|
| Maker | Moonshot | Alibaba |
| Input / 1M tokens | $3 | $1.65 |
| Output / 1M tokens | $15 | $4.95 |
| Through Somnus (at 2×) | $1.5 / $7.5 | $0.83 / $2.48 |
Cost per request
| Workload | Kimi K3 | Qwen 3.8 Max |
|---|---|---|
| Chat reply (1.5k in / 500 out) | $0.01 | $0.0050 |
| Summary of a long document (20k in / 800 out) | $0.07 | $0.04 |
| Code generation (4k in / 3k out) | $0.06 | $0.02 |
Computed from list prices. Actual token counts vary by tokenizer and prompt.
Which should you pick?
Qwen 3.8 Max is cheaper on both input and output tokens.
Price is only one factor. Try both on a sample of your own prompts — with Somnus that's the same key and the same code, only the model name changes.
Questions
Can I use Kimi K3 and Qwen 3.8 Max with one API key?
Yes. Both are available through Somnus from the same key and balance.
Are these the official prices?
They are the list prices per 1M tokens that Somnus bills usage at. Your credit is worth at least 2× what you pay.
Try both on your own prompts
Switch between Kimi K3 and Qwen 3.8 Max by changing one string. New accounts get $5 free.
Get free creditBrowse more: Alternatives · Compare · Use cases · Developers · Solutions · Tools