SettledModels

Inception releases Mercury 3, a diffusion LLM it says runs 1,000+ tokens per second

Inception says Mercury 3, a diffusion-based LLM, generates 1,068 tokens per second on Nvidia GPUs with a 1M-token context. Pricing is $0.20 input and $0.70 output per million tokens, halved at launch. It is on Inception's API, OpenRouter, Baseten and Vercel. Inception claims it beats GPT-6 Luna and Claude Haiku 5.5 but publishes no scores.

HEAT
OUTLETS: 8
FIRST SEEN: Oct 19, 2026, 22:10 ET
LAST SIGNAL: 5d ago

Primary source

Inceptioninceptionlabs.ai/blog/introducing-mercury-3

Coverage · 9 articles, oldest first

Monday, October 19, 2026

22:10
The Deep View
Inception launches Mercury 3, a diffusion-based LLM it claims is "comparable to Luna and Haiku models" but 4x to 8x faster, generating ~1,000 tokens per second
22:10
RuntimeWire
runtimewire.com/article/inception-mercury-3-diffusion-llm-launch
22:10
Unite.AI
unite.ai/inception-releases-mercury-3-its-flagship-diffusion-llm/
23:10
The Neuron
theneuron.ai/news/mercury-3-review-diffusion-language-models/

Tuesday, October 20, 2026

00:03
InceptionPrimary
Inception releases Mercury 3, a diffusion LLM it says runs 1,000+ tokens per second
01:40
Inception
inceptionlabs.ai/blog/introducing-mercury-3
09:40
Forkast
forkast.news/inceptions-mercury-3-undercuts-the-speed-optimized-frontier-at-1068-tokens-per-second/
18:45
SiliconANGLE
Inception says Mercury 3 beats GPT-6 Luna and Claude Haiku 5.5 at up to 8x the speed

Wednesday, October 21, 2026

10:45
The Decoder
Inception's Mercury 3 diffusion model claims to beat GPT-6 Luna and Claude Haiku 5.5 at a fraction of the wait

ShareX LinkedIn Hacker News Email

← The whole wire