ZMIME
Inception · Released Sep 8, 2026

Mercury 2.5

Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception.

  • Proprietary
  • Reasoning
  • Tool calling
Capability
—No ECI score yet
Input
$0.04per 1M tokens
Output
$0.15per 1M tokens
Context
260K66K max output
Pricing

What it costs

Input
$0.04per million tokens
Output
$0.15per million tokens
Cached input
$0.004per million tokens
Blended (3:1)
$0.068cheaper than 95% of priced models

Typical monthly bills

Monthly usageEstimated cost
Side project2M input + 0.5M output tokens$0.155
Team assistant25M input + 5M output tokens$1.75
Production app250M input + 50M output tokens$17.50

Official Inception API price, as listed on models.dev.

Capability

Independent benchmarks

Epoch AI has not published a Capabilities Index score for Mercury 2.5 yet.

Specifications

The details

Lab
Inception
Released
Sep 8, 2026
Knowledge cutoff
Nov 1, 2025
Context window
260,000 tokens
Max output
65,536 tokens
Inputs
Text
Output
Text
Reasoning
Adjustable effort low · medium · high
Tool calling
Yes
Structured output
Yes
Weights
Proprietary
API model ID
mercury-2.5on Inception
Availability
1 API providerlisted on models.dev

Inception model documentation

Same lab

More from Inception

Questions

About Mercury 2.5

How much does Mercury 2.5 cost?

Mercury 2.5 costs $0.04 input / $0.15 output per million tokens (official Inception API price). Cached input is $0.004 per million tokens. At a 3:1 input-to-output mix that is $0.068 per million tokens, cheaper than 95% of the 360 priced models we track.

What is the context window of Mercury 2.5?

Mercury 2.5 accepts up to 260,000 tokens per request and can write up to 65,536 tokens in one response.

How good is Mercury 2.5?

Epoch AI has not published a Capabilities Index score for Mercury 2.5 yet.

Is Mercury 2.5 open source?

No. Mercury 2.5 is proprietary; you use it through Inception’s API or partner platforms.

What inputs does Mercury 2.5 support?

Mercury 2.5 accepts text and replies in text. It is a reasoning model with low, medium and high effort settings, supports tool calling and can return structured JSON output.

When was Mercury 2.5 released?

Inception released Mercury 2.5 on Sep 8, 2026. Its training data runs to Nov 1, 2025.

What are the best alternatives to Mercury 2.5?

The closest current models from other labs on capability, price and release date are Laguna XS 2.1 (Poolside, $0.06 / $0.12), Nemotron 3.5 Lightning 30B A3B (NVIDIA, $0.05 / $0.20), Ling 3.0 Flash Fin (inclusionAI, $0.075 / $0.22) and Gemma 4 E2B IT (Google, $0.07 / $0.09).