Moonlight 16B – Efficient MoE Model for English, Code, Math
Moonlight-16B-A3B-Instruct is a Mixture-of-Experts (MoE) language model with 16 billion total parameters and 3B active per inference, designed by Moonshot AI for instruction-following, multilingual understanding, and efficient deployment. This model strikes an impressive performance-per-FLOP balance, outperforming similar-sized models like Llama 3 3B and DeepSeek-v2-Lite across benchmarks in…
3 subscribers
1 endpoints
The in-depth APIMemo review for this API hasn't been published yet —
the data below comes straight from the public marketplace listing.
Moonlight 16B – Efficient MoE Model for English, Code, Math endpoints
| Method | Endpoint | Description |
|---|---|---|
| POST |
Chat Completions /moonlight/chat |
add your prompt and interact with model |
Moonlight 16B – Efficient MoE Model for English, Code, Math pricing
| Plan | Price | Rate limit | Quotas |
|---|---|---|---|
| BASIC | Free | — |
|
| PRO | $5 / month | — |
|
| ULTRA | $15 / month | — |
|
| MEGA | $30 / month | — |
|
