llmboard.ai
Benchmarks
CompareRankings
llmboard.ai
Benchmarks
CompareRankings
HomeModelsMercury 2

Inception model product

Mercury 2

Mercury 2 is the fastest reasoning LLM, built on diffusion-based language model (dLLM) architecture. Instead of generating text token-by-token, it refines multiple text blocks simultaneously, achieving over 1,000 tokens per second on Nvidia Blackwell GPUs — 5x faster than leading speed-optimized LLMs. Supports tool usage and JSON output with 128K context window.

Updated Aug 10, 2026. Default version: Mercury 2

Compare
LLMBoard score50.8Mercury 2
Coverage60%5 benchmark families
Context window128KTokens
Official input price$0.25Inception API

On this page

  • Specification
  • Capability
  • Benchmarks
  • Arena
  • Pricing
  • Versions
  • About
  • Compare
  • Similar models
  • FAQ

Model specification

Structured fields from the published default version.

Version
Mercury 2
Released
Feb 24, 2026
Knowledge cutoff
Unknown
Parameters
N/A
Context window
128K
Max output
8.2K
Inputs
text
Outputs
text
Open weights
No
License
Proprietary

Capability profile

This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.

Mercury 2 category scores

Benchmark results

Published benchmark records for the scored version Mercury 2.

6 rows
Columns

Show columns

IFBench0.7142851.9%CAug 7, 2026
SciCode0.4151817.6%CAug 7, 2026
Tau2 Airline0.5192318.2%CAug 7, 2026
LiveCodeBench0.7247368.1%CAug 7, 2026
AIME 20250.94311462.8%CAug 7, 2026
GPQA0.711323351.7%CAug 7, 2026

Arena results

Preference and agent-evaluation signals from published Arena datasets.

47 rows
Columns

Show columns

webdevwebdev-react931155.9784N/AAug 6, 2026
webdevwebdev-html1061198.5103N/AAug 6, 2026
webdevoverall1071166.3888N/AAug 6, 2026
webdevwebdev1071166.3888N/AAug 6, 2026
text factualityenglish1381393.81,377N/AAug 6, 2026
text factualitynon english1411349.91,721N/AAug 6, 2026
text factualityoverall1441372.23,098N/AAug 6, 2026
text factualityhard prompts1441387.11,751N/AAug 6, 2026
text factualityexclude ties1451341.42,155N/AAug 6, 2026
textcoding1521393.0764N/AAug 6, 2026
texthard prompts english1601378.0855N/AAug 6, 2026
textindustry software and it services1601386.51,081N/AAug 6, 2026
textindustry medicine and healthcare1611369.9227N/AAug 6, 2026
textindustry business and management and financial operations1621351.6481N/AAug 6, 2026
textenglish1641379.21,378N/AAug 6, 2026
textexpert1651355.6229N/AAug 6, 2026
textmulti turn1681355.6536N/AAug 6, 2026
texthard prompts1701362.61,751N/AAug 6, 2026
textoverall1721357.63,099N/AAug 6, 2026
textindustry entertainment and sports and media1721318.9653N/AAug 6, 2026
textexclude ties1751315.52,156N/AAug 6, 2026
textnon english1751332.31,721N/AAug 6, 2026
textinstruction following1821322.9834N/AAug 6, 2026
text style controlcoding1871396.4764N/AAug 6, 2026
textlonger query1881327.9838N/AAug 6, 2026
textindustry life and physical and social science1911346.3469N/AAug 6, 2026
text style controlexpert1921347.4229N/AAug 6, 2026
text style controlindustry software and it services1921384.71,081N/AAug 6, 2026
text style controlindustry entertainment and sports and media1931313.2653N/AAug 6, 2026
text style controlmulti turn1951341.8536N/AAug 6, 2026
textcreative writing1961295.2524N/AAug 6, 2026
text style controlnon english1961325.01,721N/AAug 6, 2026
textindustry writing and literature and language1971303.9688N/AAug 6, 2026
text style controloverall1971346.63,099N/AAug 6, 2026
text style controlenglish1971364.81,378N/AAug 6, 2026
textrussian1981303.8358N/AAug 6, 2026
text style controlhard prompts english1981372.1855N/AAug 6, 2026
text style controlhard prompts1991359.81,751N/AAug 6, 2026
text style controlindustry business and management and financial operations2021340.7481N/AAug 6, 2026
text style controlexclude ties2031299.52,156N/AAug 6, 2026
text style controlinstruction following2051322.9834N/AAug 6, 2026
text style controlindustry medicine and healthcare2071345.3227N/AAug 6, 2026
text style controlcreative writing2151299.3524N/AAug 6, 2026
text style controlindustry writing and literature and language2151310.1688N/AAug 6, 2026
text style controlrussian2261304.9358N/AAug 6, 2026
text style controllonger query2341323.6838N/AAug 6, 2026
text style controlindustry life and physical and social science2461323.4469N/AAug 6, 2026

Pricing

Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.

Official API
$0.25 input, $0.75 output per 1M
Official provider
Inception
Lowest third-party
From $0.25 input, $0.75 output per 1M via NanoGPT
Tracked offerings
5
5 rows
Columns

Show columns

NanoGPTmercury-2global$0.25$0.75128KAug 7, 2026
Inceptionmercury-2global$0.25$0.75128KAug 7, 2026
Vercel AI Gatewayinception/mercury-2global$0.25$0.75128KAug 7, 2026
OpenRouterinception/mercury-2global$0.25$0.75128KAug 7, 2026
Venice AImercury-2global$0.3125$0.9375128KAug 7, 2026

Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.

Mercury 2 versions

All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.

1 rows
Columns

Show columns

Mercury 2Feb 24, 202650.8N/A128K8.2KNoProprietary

What is Mercury 2?

A concise description based on the published model registry.

Mercury 2 is the fastest reasoning LLM, built on diffusion-based language model (dLLM) architecture. Instead of generating text token-by-token, it refines multiple text blocks simultaneously, achieving over 1,000 tokens per second on Nvidia Blackwell GPUs — 5x faster than leading speed-optimized LLMs. Supports tool usage and JSON output with 128K context window.

Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.

Data snapshot: 2026-08-07. Editorial model content is not available in the backend.

Mercury 2 vs nearby models

Open a comparison with the three ranked models immediately above and below this model.

Mercury 2vsQwen2.5 VL 32BMercury 2vsKimi K2Mercury 2vsGPT-4-TurboMercury 2vsPhi 4 Reasoning PlusMercury 2vsNova LiteMercury 2vsMistral Small 4

Models similar to Mercury 2

Recommendations prioritize the same model type and family, then the closest published LLMBoard score.

#81+0.3
OP

GPT-4-Turbo

OpenAI

51.1 LLMBoard

DetailsCompare
#83-0.5
MI

Phi 4 Reasoning Plus

Microsoft

50.3 LLMBoard

DetailsCompare
#80+0.6
MA

Kimi K2

Moonshot AI

51.4 LLMBoard

DetailsCompare
#79+0.7
AC

Qwen2.5 VL 32B

Alibaba Cloud / Qwen Team

51.5 LLMBoard

DetailsCompare
#78+0.9
AC

Qwen3 VL 32B Thinking

Alibaba Cloud / Qwen Team

51.7 LLMBoard

DetailsCompare
#77+1.6
ME

Llama 3.1 70B

Meta

52.4 LLMBoard

DetailsCompare

FAQ

Common questions about Mercury 2.

When was Mercury 2 released?

Mercury 2's default version was released on Feb 24, 2026.

How much does Mercury 2 cost?

Mercury 2's official API price is $0.25 per million input tokens and $0.75 per million output tokens via Inception. The lowest tracked third-party offer starts at $0.25 input and $0.75 output via NanoGPT.

Who created Mercury 2?

Mercury 2 is published under Inception in the model registry.

What is the context window for Mercury 2?

The default version has a 128K token context window.

Is Mercury 2 open weight?

No. The default version is not marked as having publicly available weights.

How many API providers offer Mercury 2?

5 published provider offerings are linked to the default version.

What models should I compare Mercury 2 with?

Nearby ranked alternatives include Qwen2.5 VL 32B, Kimi K2, GPT-4-Turbo.

Rankings

OverallCodingText ArenaPricing

Modalities

Image GenerationVideo GenerationSpeech-to-TextEmbeddings

Benchmarks

All BenchmarksReasoningMathCoding

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai