MiniMax · 2026-06 · 稀疏 MoE · 多模態
MiniMax-M3 MXFP8
MiniMax-M3 MXFP8 是 MiniMax 於 2026-06 發布的 分組查詢注意力（GQA） Transformer，共 60 層，隱藏維度 6144，上下文視窗 1,048,576 個 token。
層堆疊
注意力 ×60 · 層數 60
關鍵參數
| 實驗室 | MiniMax |
|---|---|
| 發布時間 | 2026-06 |
| 總參數 | 443.4 B |
| 激活參數 | 25.5 B |
| 上下文 | 1,048,576 tokens |
| 注意力 | GQA (64:4) |
| 層數 | 60 |
| 隱藏維度 | 6,144 |
| 注意力頭 | 64 |
| 詞表 | 200,064 |
| 位置編碼 | RoPE |
| 正規化 | RMSNorm |
| 激活函數 | swigluoai |
| 訓練精度 | MXFP8 |
| 架構類別 | MiniMaxM3SparseForConditionalGeneration |
架構概覽
注意力
分組查詢注意力（GQA） — 64 q-heads / 4 kv-heads. RoPE θ=5,000,000.
前饋 / MoE
它是稀疏混合專家模型：128 個路由專家、top-4 路由；每個 token 約激活 25.5B 參數。
設定中包含 1 個額外的多 token 預測（MTP）層，用於加速解碼。
欄位級對比
對比對象為同實驗室的上一代模型；無前代時取結構最相似的模型。倍數欄 = 本模型數值 ÷ 對比模型數值。
與前代 MiniMax-M3 的全欄位對比
| 模型 | MiniMax-M3 MXFP8 | MiniMax-M3 | 倍數 |
|---|---|---|---|
| model_type | minimax_m3_vl | minimax_m3_vl | ＝ |
| architectures | MiniMaxM3SparseForConditionalGeneration | MiniMaxM3SparseForConditionalGeneration | ＝ |
| hidden_size | 6144 | 6144 | ≈1 |
| num_hidden_layers | 60 | 60 | ≈1 |
| num_attention_heads | 64 | 64 | ≈1 |
| num_key_value_heads | 4 | 4 | ≈1 |
| head_dim | 128 | 128 | ≈1 |
| intermediate_size | 3072 | 3072 | ≈1 |
| hidden_act | swigluoai | swigluoai | ＝ |
| num_experts_per_tok | 4 | 4 | ≈1 |
| n_shared_experts | 1 | 1 | ≈1 |
| rope_theta | 5000000 | 5000000 | ≈1 |
| max_position_embeddings | 1048576 | 1048576 | ≈1 |
| rms_norm_eps | 0.000001 | 0.000001 | ≈1 |
| vocab_size | 200064 | 200064 | ≈1 |
| torch_dtype | bfloat16 | bfloat16 | ＝ |
| num_nextn_predict_layers | — | 1 | |
| image_grid_pinpoints | [(336, 336), (336, 672), (336, 1008), (336, 1344), (336, 1680), (336, 2016), (672, 336), (672, 672), (672, 1008), (672, 1344), (672, 1680), (672, 2016), (1008, 336), (1008, 672), (1008, 1008), (1008, 1344), (1008, 1680), (1008, 2016), (1344, 336), (1344, 672), (1344, 1008), (1344, 1344), (1344, 1680), (1344, 2016), (1680, 336), (1680, 672), (1680, 1008), (1680, 1344), (1680, 1680), (1680, 2016), (2016, 336), (2016, 672), (2016, 1008), (2016, 1344), (2016, 1680), (2016, 2016)] | [(336, 336), (336, 672), (336, 1008), (336, 1344), (336, 1680), (336, 2016), (672, 336), (672, 672), (672, 1008), (672, 1344), (672, 1680), (672, 2016), (1008, 336), (1008, 672), (1008, 1008), (1008, 1344), (1008, 1680), (1008, 2016), (1344, 336), (1344, 672), (1344, 1008), (1344, 1344), (1344, 1680), (1344, 2016), (1680, 336), (1680, 672), (1680, 1008), (1680, 1344), (1680, 1680), (1680, 2016), (2016, 336), (2016, 672), (2016, 1008), (2016, 1344), (2016, 1680), (2016, 2016)] | ＝ |
| image_seq_length | 576 | 576 | ≈1 |
| image_token_index | 200025 | 200025 | ≈1 |
| video_token_index | 200026 | 200026 | ≈1 |
| multimodal_projector_bias | true | true | ＝ |
| num_reward_heads | 0 | 0 | ＝ |
| process_image_mode | dynamic_res | dynamic_res | ＝ |
| projector_hidden_act | gelu | gelu | ＝ |
| vision_feature_layer | -1 | -1 | ≈1 |
結構最相似的模型
- MiniMax-M3 (MiniMax, 2026-06) 1.000
- Qwen3.8 2.4T A95B (Qwen, 2026-08) 0.927
- MiniMax-Text-01 (MiniMax, 2025-01) 0.926
- MiniMax-M1 (MiniMax, 2025-06) 0.926
- GLM-4.6 (GLM, 2025-10) 0.924
- GLM-4.5 (GLM, 2025-07) 0.922
相似度取值範圍 0（無共同類別特徵）到 1（結構輪廓完全一致）。
原始 config 欄位
46 個欄位
| architectures | MiniMaxM3SparseForConditionalGeneration |
|---|---|
| model_type | minimax_m3_vl |
| image_grid_pinpoints | [(336, 336), (336, 672), (336, 1008), (336, 1344), (336, 1680), (336, 2016), (672, 336), (672, 672), (672, 1008), (672, 1344), (672, 1680), (672, 2016), (1008, 336), (1008, 672), (1008, 1008), (1008, 1344), (1008, 1680), (1008, 2016), (1344, 336), (1344, 672), (1344, 1008), (1344, 1344), (1344, 1680), (1344, 2016), (1680, 336), (1680, 672), (1680, 1008), (1680, 1344), (1680, 1680), (1680, 2016), (2016, 336), (2016, 672), (2016, 1008), (2016, 1344), (2016, 1680), (2016, 2016)] |
| image_seq_length | 576 |
| image_token_index | 200025 |
| video_token_index | 200026 |
| multimodal_projector_bias | true |
| num_reward_heads | 0 |
| process_image_mode | dynamic_res |
| projector_hidden_act | gelu |
| vision_feature_layer | -1 |
| vision_feature_select_strategy | full |
| torch_dtype | bfloat16 |
| transformers_version | 4.52.4 |
| projector_hidden_size | 6144 |
| dtype | bfloat16 |
| hidden_size | 6144 |
| intermediate_size | 3072 |
| num_hidden_layers | 60 |
| num_attention_heads | 64 |
| num_key_value_heads | 4 |
| head_dim | 128 |
| vocab_size | 200064 |
| max_position_embeddings | 1048576 |
| rms_norm_eps | 0.000001 |
| use_gemma_norm | true |
| attention_output_gate | false |
| rope_theta | 5000000 |
| rotary_dim | 64 |
| partial_rotary_factor | 0.5 |
| hidden_act | swigluoai |
| use_qk_norm | true |
| tie_word_embeddings | false |
| dense_intermediate_size | 12288 |
| shared_intermediate_size | 3072 |
| num_local_experts | 128 |
| num_experts_per_tok | 4 |
| n_shared_experts | 1 |
| scoring_func | sigmoid |
| use_routing_bias | true |
| moe_layer_freq | 0×3 + 1×57 |
| qk_norm_type | per_head |
| num_mtp_modules | 1 |
| swiglu_alpha | 1.702 |
| swiglu_limit | 7 |
| routed_scaling_factor | 2 |