Mercury 2.5
Parameters
Undisclosed
Context Length
260K
Modality
Text
Architecture
Undisclosed
License
Proprietary
Release Date
8 Sept 2026
Knowledge Cutoff
-
API Pricing (per 1M)
Input: $0.04 · Output: $0.15
Evaluation Benchmarks
No evaluation benchmarks for Mercury 2.5 available.
Rankings
Overall Rank
-
Coding Rank
-
About Mercury 2.5
Mercury 2.5 is a diffusion-based language model by Inception designed for ultra-fast reasoning and text generation. By generating and refining multiple tokens in parallel, it achieves high throughput for complex problem-solving and agentic workflows.
Technical Specifications
Architecture specifications are undisclosed for proprietary models.
Attention
Attention Structure
Undisclosed
Attention Heads
Undisclosed
Key-Value Heads
Undisclosed
Attention Head Dimension
Undisclosed
Position Embedding
Undisclosed
RoPE Theta
Undisclosed
Sliding Window Attention
Undisclosed
Sliding Window Size
Undisclosed
Sliding Window Ratio
Undisclosed
Linear Attention
Undisclosed
Linear Attention Ratio
Undisclosed
Normalization
Undisclosed
Activation Function
Undisclosed
Dimensions
Hidden Dimension Size
Undisclosed
Number of Layers
Undisclosed
FFN Intermediate Size (Dense)
Undisclosed
Multi-Token Prediction Heads
Undisclosed
Tokenizer
Vocabulary Size
Undisclosed
About Mercury
The Mercury model family developed by Inception.