Loading page
FlagshipInception
Mercury 2
Mercury 2 is Inception's flagship text-only model with a 128K token context window, designed for high-throughput text generation tasks.
Context 128K
Tier Flagship
Input from
$0.250 / 1M tokens
across 1 provider
API Pricing
| Provider | Input / 1M | Output / 1M | Cached / 1M |
|---|---|---|---|
| $0.250 | $0.750 | $0.025 |
Prices updated daily. Last check: Sep 6, 2026
Mercury 2 pricing by provider
Input, output, and batch rates, plus alternatives, for one provider at a time.
Performance & Benchmarks
Source: Artificial Analysis →Intelligence
14.9 / 100
Coding
31.1 / 100
Output Speed
795 t/s
Latency (TTFT)
3.9s
Reasoning & Knowledge
- GPQA Diamond77.0%
- Humanity's Last Exam17.1%
Agentic & Tool Use
- Terminal-Bench Hard26.5%
- Terminal-Bench v2.127.3%
- τ²-bench70.8%
- τ-bench Banking9.5%
Instruction & Long Context
- IFBench69.8%
- Long-Context Reasoning43.7%
Benchmarks measured Sep 2026. Scores are independent evaluations, not vendor-reported.
Model Details
General
- Creator
- Inception
- Family
- Mercury
- Tier
- Flagship
- Context Window
- 128K
- Modalities
- Text
Capabilities
- Tool Calling
- No
- Open Source
- No
Strengths & Limitations
Strengths
- High output token generation speed at 918.2 tokens per second
- Large 128K token context window for processing lengthy documents
- Flagship-tier model representing Inception's most capable offering
- Text-focused design optimized for language tasks
- Substantial context capacity for document analysis and long-form generation
Limitations
- No function calling or tool use capabilities
- Text-only modality with no image or multimodal support
- High time to first token at 3.6 seconds affects interactive use cases
- Proprietary model with no open-source availability
- Smaller context window compared to some frontier models with 200K+ tokens
Key Features
•128K token context window
•Text input and output processing
•High-speed token generation (918+ tokens/second)
•Flagship-tier language understanding
•Document processing capabilities
•Long-form content generation
•Streaming text responses
About Mercury 2
Mercury 2 is Inception's flagship model in the Mercury family, representing the company's top-tier offering for text-based AI applications. As a proprietary model, it focuses exclusively on text processing and generation without multimodal capabilities.
The model features a 128K token context window and demonstrates strong performance characteristics, with benchmark data showing output generation speeds of over 918 tokens per second. However, it has a relatively high time to first token at approximately 3.6 seconds and does not support function calling or tool use capabilities.
Mercury 2 is positioned for applications requiring high-volume text processing where throughput is prioritized over interactive response times. Its substantial context window makes it suitable for document analysis and long-form content generation, though users requiring multimodal capabilities or tool integration would need to consider alternative models.
Common Use Cases
Mercury 2 is well-suited for high-volume text processing applications where throughput takes priority over immediate response times. Its 128K context window and fast generation speed make it effective for batch document analysis, content creation workflows, summarization of lengthy texts, and automated writing tasks. The model works best in scenarios where the 3.6-second initial response delay is acceptable, such as background processing jobs, content pipeline automation, or applications where users expect to wait for comprehensive outputs rather than interactive conversations.
Frequently Asked Questions
How much does Mercury 2 cost per million tokens?
Mercury 2 pricing varies by provider and usage patterns. Check the pricing table above for current rates across all available providers offering this model.
What is Mercury 2 best used for?
Mercury 2 excels at high-throughput text processing tasks like batch document analysis, content generation, and long-form writing where its 918+ tokens/second generation speed and 128K context window provide advantages. It's ideal for non-interactive workflows where the 3.6-second startup time is acceptable.
Does Mercury 2 support function calling or multimodal inputs?
No, Mercury 2 is a text-only model without function calling capabilities or support for images or other modalities. It focuses exclusively on text input and output processing.