Kimi · 2025-02 · Sparse MoE
Moonlight 16B
Moonlight 16B is a multi-head latent attention (MLA) transformer released by Kimi in 2025-02, with 27 layers, hidden size 2048 and a context window of 8,192 tokens.
Layer stack
Attention ×27 · Layers 27
Key facts
| Family | Kimi |
|---|---|
| Released | 2025-02 |
| Params | 15.8 B |
| Active | 2.8 B |
| Context | 8,192 tokens |
| Attention | MLA · kv_lora_rank 512 |
| Layers | 27 |
| Hidden | 2,048 |
| Heads | 16 |
| Vocab | 163,840 |
| Position | RoPE |
| Norm | RMSNorm |
| Activation | silu |
| Dtype | bf16 |
| Architecture class | DeepseekV3ForCausalLM |
Architecture overview
Attention
multi-head latent attention (MLA). kv_lora_rank=512. RoPE θ=50,000.
Feed-forward / MoE
It is a sparse mixture-of-experts with 64 routed experts and top-6 routing; about 2.8B parameters are active per token.
Field-level comparison
Compared against the previous model of the same lab; where none exists, against the structurally closest model. The ratio column is this model divided by the comparison model.
Full-field comparison vs most-similar DeepSeek-V2 Lite
| Model | Moonlight 16B | DeepSeek-V2 Lite | Ratio |
|---|---|---|---|
| model_type | deepseek_v3 | deepseek_v2 | ≠ |
| architectures | DeepseekV3ForCausalLM | DeepseekV2ForCausalLM | ≠ |
| hidden_size | 2048 | 2048 | ≈1 |
| num_hidden_layers | 27 | 27 | ≈1 |
| num_attention_heads | 16 | 16 | ≈1 |
| num_key_value_heads | 16 | 16 | ≈1 |
| intermediate_size | 11264 | 10944 | ×1.03 |
| hidden_act | silu | silu | ＝ |
| n_routed_experts | 64 | 64 | ≈1 |
| num_experts_per_tok | 6 | 6 | ≈1 |
| n_shared_experts | 2 | 2 | ≈1 |
| moe_intermediate_size | 1408 | 1408 | ≈1 |
| kv_lora_rank | 512 | 512 | ≈1 |
| qk_rope_head_dim | 64 | 64 | ≈1 |
| rope_theta | 50000 | 10000 | ×5.00 |
| max_position_embeddings | 8192 | 163840 | ×0.050 |
| rms_norm_eps | 0.00001 | 0.000001 | ×10.00 |
| vocab_size | 163840 | 102400 | ×1.60 |
| torch_dtype | bfloat16 | bfloat16 | ＝ |
| num_nextn_predict_layers | 0 | — | |
| attention_bias | false | false | ＝ |
| attention_dropout | 0 | 0 | ＝ |
| aux_loss_alpha | 0.001 | 0.001 | ≈1 |
| bos_token_id | 163584 | 100000 | ×1.64 |
| eos_token_id | 163585 | 100001 | ×1.64 |
| ep_size | 1 | — |
Most similar architectures
- DeepSeek-V2 Lite (DeepSeek, 2024-05) 0.997
- GLM-4.7-Flash (GLM, 2026-01) 0.994
- Kimi Linear 48B A3B (Kimi, 2025-10) 0.992
- DeepSeek-V2 (DeepSeek, 2024-05) 0.983
- DeepSeek-V3 (DeepSeek, 2024-12) 0.913
- DeepSeek-V3 Base (DeepSeek, 2024-12) 0.913
Similarity values range from 0 (no shared categorical features) to 1 (identical profiles).
Raw config fields
42 fields
| architectures | DeepseekV3ForCausalLM |
|---|---|
| attention_bias | false |
| attention_dropout | 0 |
| aux_loss_alpha | 0.001 |
| bos_token_id | 163584 |
| eos_token_id | 163585 |
| ep_size | 1 |
| first_k_dense_replace | 1 |
| hidden_act | silu |
| hidden_size | 2048 |
| initializer_range | 0.02 |
| intermediate_size | 11264 |
| kv_lora_rank | 512 |
| max_position_embeddings | 8192 |
| model_type | deepseek_v3 |
| moe_intermediate_size | 1408 |
| moe_layer_freq | 1 |
| n_group | 1 |
| n_routed_experts | 64 |
| n_shared_experts | 2 |
| norm_topk_prob | true |
| num_attention_heads | 16 |
| num_experts_per_tok | 6 |
| num_hidden_layers | 27 |
| num_key_value_heads | 16 |
| num_nextn_predict_layers | 0 |
| pretraining_tp | 1 |
| qk_nope_head_dim | 128 |
| qk_rope_head_dim | 64 |
| rms_norm_eps | 0.00001 |
| rope_theta | 50000 |
| routed_scaling_factor | 2.446 |
| scoring_func | sigmoid |
| seq_aux | true |
| tie_word_embeddings | false |
| topk_group | 1 |
| topk_method | noaux_tc |
| torch_dtype | bfloat16 |
| transformers_version | 4.46.3 |
| use_cache | true |
| v_head_dim | 128 |
| vocab_size | 163840 |