Inception

Congrats to the @_inception_ai team on launching Mercury Decide, the highest-accuracy decision model at the lowest cost!

Try it from the Baseten Model Library: baseten.co/library/mercur…

One week ago we launched Mercury Decide, @_inception_ai System One decision model. It hit #2 on OpenRouter by users, and we've been improving it every day since.

Independent results from @ValsAI show that Mercury Decide is more accurate and more affordable than Jev.
Proud of how fast this team moves.

Free on @OpenRouter and available through the Inception API and @baseten. I'd love to see what you build.

openrouter.ai/inception/merc…

Mercury Decide from @_inception_ai is now available with zero data retention on OpenRouter

The fastest growing decision model on OpenRouter last week now has a paid ZDR endpoint, alongside the free one

$0.02/M input (50% off the $0.04 list price). Output and cached input are free. 66K context

openrouter.ai/inception/merc…

Last week, we launched Mercury Decide, a Jev-beating decision model that quickly rose to #2 on OpenRouter by number of users: openrouter.ai/inception/merc…

This week, we’re updating Mercury Decide with substantially better benchmark performance, longer context (33k → 66k), General Availability with Zero Data Retention (ZDR), and more.

All that with lower pricing than Jev (ZDR for $0.02/$0 per 1M in/out, discounted at 50% from the list price of $0.04/$0, plus 100M free tokens), alongside an unlimited free tier.

What have we improved? Read on to find out! [🧵1/4]

We're excited to welcome Michael Giannangeli (@mrgiannangeli) to Inception as our new Head of Product! 📷

Michael has spent years building AI products that millions of people use every day. At Amazon, he led product and agentic AI for Nova foundation models, building the capabilities for intelligent systems to reason, act, and scale in the real world. Before that, Michael ran the Echo and Echo Dot product lines, started Alexa’s computer vision team, and helped scale Alexa from 1M to 100M+ customers.

He holds an MBA from MIT. Before tech, he worked as an economic consultant, did stints on Wall Street and Capitol Hill, and played NCAA basketball.

At Inception, Michael will lead product as we bring Mercury, our family of diffusion LLMs, to more developers and enterprises. Mercury generates tokens in parallel rather than one at a time, so it's much faster and cheaper than standard autoregressive models. Michael knows how to turn breakthrough AI into products people love, which makes him a great fit for this next phase.

Welcome to the team, Michael! 🎉

🤖 Last week, I compiled an awesome list of Awesome Jev lists and found the most widespread applications for decision models were to be found in AI coding harnesses like Claude Code and OpenCode.

For example, one challenging task for coding agents is to detect Observation Prompt Injection (OPI). I compared a handful of decision models to see how they stack up, results in the table.

What is Observation Prompt Injection (OPI)? Let's dig a bit deeper! (🧵1/4)

Inception is heading to #COLM2026!

@StefanoErmon and @adityagrover_ will be speaking at the Non-Autoregressive Language Models workshop on Oct. 9.

We’re also hosting an Inception social on Oct. 8! We’re almost at capacity, but have a few spots left. RSVP approval is required.

Event link: luma.com/597gw374

This week we launched Mercury Voice, a dLLM optimized for voice agents with 2x+ lower latency than models like GPT-6 Luna.

In this dental clinic receptionist demo, median end-to-end latency stays under 300ms, tool calls included.
See (or hear?) it yourself in the model playground: chat.inceptionlabs.ai/voice

Enterprise customers interested in Mercury Voice can contact us at sales@inceptionlabs.ai to get access.

What can you build with Mercury Decide, our new decision model, free for early access on @OpenRouter?

Our Developer Relations Engineer, @stefan_webb, has compiled a resource analyzing the breadth of applications across the web: github.com/stefanwebb/met…

The #1 application has been coding agents: compaction, routing, guardrails, and verification.

Learn more, including insights into how to best use decision models at the link above.

Model access details here: openrouter.ai/inception/merc…

Introducing Mercury Decide: the most intelligent decision model on @OpenRouter (JevBench v1.4) and one of the fastest.

Up to 14 decisions per second. Watch it go head-to-head with Jev in chess.

Free for early access on OpenRouter: openrouter.ai/inception/merc…

Today, we’re introducing Mercury Voice, a diffusion LLM specialized for agentic voice applications.

Mercury Voice delivers 2x+ lower latency than models including GPT-6 Luna, Gemma 4 31B, and Claude Haiku 4.5, while beating them on a range of voice benchmarks.

Enterprise customers interested in Mercury Voice can contact us at sales@inceptionlabs.ai to get access.

The right way to measure AI value: intelligence per dollar.

Our co-founder and CEO @StefanoErmon joined @YahooFinance Live to talk about Mercury 2.5, our newest and most capable diffusion LLM model yet, and why he thinks about it this way. Thanks for having him @BrookeDiPalma!

Mercury powers real-time voice systems with the tightest latency budgets.

@OpenCall_AI runs AI phone agents on live patient calls.

After moving off a specialized chip provider, they scaled from 300M to over 1B tokens per day, with median model response latency under 200ms.

Today, we’re introducing Mercury 2.5, the most capable diffusion LLM on the market.

It offers a 40% jump in intelligence over Mercury 2, and runs over 1,100 tokens/sec on widely-available @NVIDIAAI GPUs.

It’s available today on our API, @OpenRouter, and @Baseten.

Contact us to evaluate Mercury 2.5 for production:
inceptionlabs.ai/models

Mercury 2.5 Preview is live exclusively on OpenRouter.

80% off pricing through 9/7 🚀

🔔New cookbook: build a customer support voice agent with Mercury and @ElevenLabs.

@anemenemily's walkthrough below: tool calls into a live database and streaming voice, all in one notebook.

With Mercury, you don't have to choose between latency and intelligence. Diffusion LLMs give you both.

Cookbook: docs.inceptionlabs.ai/cookbooks/elev…

Tokenmaxxing is out. Valuemaxxing is in says @adityagrover_ live at the Ai4 Podcast Pavilion, on the Inside @AsembleAI Podcast with Sam Dey. ⚡

The industry spent years optimizing for quality at any cost. Aditya's POV: that era's ending, and what comes next gets measured in value delivered, not tokens burned.

Listen to the full conversation: open.spotify.com/episode/3YjFbe…

Diffusion LLMs are the electric car of language models. Same interface. Fundamentally different machine under the hood.

Our VP of Engineering @kumarc1 breaks it down on @TheInfraPod, including why voice and search agents can't wait for one-token-at-a-time generation.

open.spotify.com/episode/4tGOL0…

Heading to Ray Summit this week?

Our Co-Founder and CTO, @adityagrover_ will be speaking 8/25 at 5 PM PT on Diffusion Large Language Models: Redefining the Token Efficiency Frontier.

Afterward, we’re hosting an Inception social nearby on 8/25 at 7:30 PM! Come meet the team, hang out with us, and chat about diffusion LLMs.

Space is limited, so RSVP here: luma.com/gyy4s6mq

Speed matters when every millisecond shapes the user experience. ⚡

@OpenCall_AI moved its voice agent workloads to Mercury 2, and AI search agents are running it at every step of their pipeline.

@StefanoErmon breaks down the use cases on @DellTech & @NVIDIA's Reshaping Workflows podcast: reshaping-workflows.simplecast.com/episodes/10x-f…