Mercury 2.5
Mercury 2.5Mercury 2.5 is the most capable diffusion LLM on the market.
Model details
Mercury 2.5 is the successor to Mercury 2 and Inception’s most intelligent diffusion-based reasoning model, delivering stronger reasoning and coding performance without sacrificing speed.
Instead of decoding one token at a time like standard autoregressive LLMs, it generates responses through parallel refinement, iteratively improving multiple tokens at once rather than following a long left-to-right chain.
Compared with Mercury 2, Mercury 2.5 offers:
Stronger coding, instruction-following, mathematical reasoning, and knowledge recall
A 260K-token context window, up from 128K
Lower input pricing, while retaining reasoning, native tool use, and structured output
Better support for complex agentic workflows, including agents and subagents
Mercury 2.5 is OpenAI API compatible, so it can drop into an existing stack without a rewrite. It’s best suited for complex, latency-sensitive workloads where quality and speed both matter, including rapid coding iteration, agentic systems, customer support, and enterprise search.
Inception Labs: Introducing Mercury 2.5
Here's an example calling it via the API:
1from inceptionai import Inception
2
3client = Inception() # reads INCEPTION_API_KEY from the environment
4
5completion = client.chat.completions.create(
6 model="mercury-2.5",
7 messages=[{"role": "user", "content": "What is a diffusion model?"}],
8 reasoning_effort="medium",
9 max_completion_tokens=8192,
10)
11print(completion.choices[0].message.content)1{
2 "id": "chatcmpl-7a2b3c4d5e",
3 "object": "chat.completion",
4 "created": 1745798400,
5 "model": "mercury-2.5",
6 "choices": [
7 {
8 "index": 0,
9 "finish_reason": "stop",
10 "message": {
11 "role": "assistant",
12 "content": "A diffusion language model is a type of language model that uses diffusion to generate text."
13 }
14 }
15 ],
16 "usage": {
17 "prompt_tokens": 12,
18 "completion_tokens": 8,
19 "total_tokens": 20,
20 "reasoning_tokens": 0,
21 "cached_input_tokens": 0
22 }
23}
