Mercury 2.5
Inception · released Sep 8, 2026
Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception. Instead of generating tokens sequentially, Mercury 2.5 produces and refines multiple tokens in parallel, achieving...
Specification
- Context window
- 260K
- Max output
- 66K
- Knowledge cutoff
- Not stated
- Parameters
- Undisclosed
- Licence
- Proprietary
- Serving providers
- 1
- Moderated
- No
- Uptime
- 100.0%
Intelligence
12.3
50th percentile
Coding
—
Coding Index
Agentic
—
Agentic Index
Output speed
162 t/s
Median across providers
Latency
1.47s
Time to first token
Cost per task
$0.01
Estimated
Benchmarks
Where the score comes from
The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.
Evaluation scores
Percentage correct · higher is better
- AA-LCR (long context)71.7%
- SciCode38.5%
- Humanity's Last Exam11.8%
An evaluation missing from this list was not run for this model — it is not a zero.
View as table
| Evaluation | Score |
|---|---|
| AA-LCR (long context) | 71.7% |
| SciCode | 38.5% |
| Humanity's Last Exam | 11.8% |
Against its peers
Intelligence Index · this model highlighted, nearest peers in grey
- Nemotron 3.5 LightningNVIDIA12.9
- GLM 4.5Z.ai12.8
- Kimi K2 0711Moonshot AI12.7
- GPT-4.1OpenAI12.7
- o3 MiniOpenAI12.5
- o1-proOpenAI12.4
- Mercury 2.5Inception12.3
- Sonar Reasoning ProPerplexity11.8
Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.
View as table
| Model | Intelligence |
|---|---|
| Nemotron 3.5 Lightning | 12.9 |
| GLM 4.5 | 12.8 |
| Kimi K2 0711 | 12.7 |
| GPT-4.1 | 12.7 |
| o3 Mini | 12.5 |
| o1-pro | 12.4 |
| Mercury 2.5 | 12.3 |
| Sonar Reasoning Pro | 11.8 |
Percentile among all indexed models
Pricing
What it costs to run
List prices per million tokens, plus what one representative task works out to.
List price
- Input / 1M tokens
- $0.04
- Output / 1M tokens
- $0.15
- Cached input / 1M
- $0.004
- Blended 3:1
- $0.068
One task, estimated
$0.01
- Input tokens
- 50,000
- Output tokens
- 80,000
- Profile
- Reasoning
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).
Serving providers
Speed and latency figures are medians across these providers, so a widely-served model reports a blend rather than any single endpoint.