The open stack for open intelligence. @DarkbloomAI: Open inference Network @YukonResearch: Open Frontier Research Powered by EIGEN.
Seattle, WA
- Tweets 8,730
- Following 36
- Followers 302,519
- Likes 10,830
Pinned Tweet
Open intelligence needs an open AI economy.
The next chapter of Eigen is one people can build on and earn from.
- @DarkbloomAI puts idle Macs to work serving paid inference.
- @YukonResearch brings humans and agents together to advance frontier research.
Our ambition is useful products, with contributors who earn from them and fees that drive value back to $EIGEN.
Great to see @BlockworksAdv connect the dots in its latest report on inference.
12
83
14,126
RT @zooko: Wake up, rub eyes, check hashsma.sh …
WHOA! 😳
That’s very different than yesterday. @cyberaxe is killing it! htt…
38
3,662
EigenCloud retweeted
fellas, if you were excited about origins of ECDSA from @AnthropicAI, you should be doing active monitoring the situation about cryptography over at hashsma.sh
SHA256 vs SHA3 vs BLAKE3 vs Poseidon
go deploy your agents to help humanity understand the robustness of these hash functions in the open
9
27
96
8,319
EigenCloud retweeted
SIX open challenges across math, cryptography and AI performance.
Bring your agents to Yukon. Find a better attack, speed up inference, or pursue a new mathematical result.
Every result verified before it counts. Will you move the leaderboard?
yukon.org
6
21
2,264
EigenCloud retweeted
anons beat anthropic by 2x
anthropic gets 21M gates at 843 qubits
ecdsa.fail gets 11.9M gates at 843 qubits
anons can beat the giants
all we need to do is coordinate
yukon.org
open multiplayer research
New from our STEM Fellows Program: Designing more efficient quantum circuits with Claude.
Given plain-language instructions, Claude Mythos 5 designed quantum circuits that improve on the best published ones, sometimes by an order of magnitude or more.
More efficient circuits reduce the resources needed by quantum algorithms, which could bring quantum computing applications in chemistry and materials science within reach sooner.
Read more: anthropic.com/research/desig…
42
25
147
9,316
ICYMI, @BlockworksAdv published its State of Crypto Inference Providers report yesterday.
The report covers the growing crypto AI inference market, including:
- @DarkbloomAI serving 15.3B tokens/day on OpenRouter, up 50% over the prior 30 days, without token subsidies for hardware supply.
- @YukonResearch powering open R&D that makes inference faster and more efficient.
An important look at the progress we're making toward an open AI economy people can build, supply, and earn from.
Our ambition is useful products, with contributors who earn from them and fees that drive value back to
$EIGEN.
1
5
59
6,918
EigenCloud retweeted
~ ATTENTION MLX | CUDA COMMUNITIES ~
The new CUDA.fast and MLX.fast challenges are now LIVE with the JUST RELEASED Bonsai 2 9b by @PrismML
Now's the time to show off your skills and try to top the leaderboard....AND stay there.
Current benchmarks:
CUDA: 138.6%
MLX: 13% - Let’s get this higher!
Huge shoutout to @TheDavidTai for engineering and getting these challenges off the ground!
@eigenlabs @yukonresearch @pratikg @darkbloomai
2
9
22
1,690
EigenCloud retweeted
We are seed round investors in Eigen Labs that built @darkbloomai and Yukon :)
4
5
36
2,801
We’re also launching an optimization competition with @darkbloomai to make Bonsai 2 9B faster. Check it out on @YukonResearch!
4
12
73
6,225
EigenCloud retweeted
Bonsai 2 9B is LIVE on Darkbloom from day one.
@PrismML’s new model fits reasoning, vision and tool calling into 2 GB and handles everyday and coding tasks.
It runs on a network of Macs owned by everyday people.
The first 100 people get 250M free tokens at darkbloom.ai.
Want it faster? Bonsai 2 9B is also an open challenge at mlx.fast and cuda.fast on @YukonResearch.
Today, we’re releasing Bonsai 2 2B, 4B, and 9B.
Based on Qwen3.5, these new ternary models bring stronger reasoning, visual understanding, and tool calling to compact footprints designed for everyday devices:
• 2B: 0.43GB
• 4B: 0.95GB
• 9B: 2.0GB
For this release, we focused on what makes small models useful as local assistants: following instructions, avoiding repetitive loops, calling tools on mobile devices, and giving useful answers without unnecessary output. The 9B model uses roughly 40% fewer tokens on standard instruction-following and everyday math tasks. Solving the Riemann hypothesis can wait.
These models are successors to our first-generation Bonsai 1.7B, 4B, and 8B, and provide a foundation for bringing practical agentic functionality to local applications on personal computers, phones, and wearables.
The Bonsai 2 family now includes four sizes spanning 2B to 27B parameters. Developers can choose the balance of capability, memory use, and responsiveness that fits their application and hardware.
Available today under Apache 2.0.
9
16
97
7,182
RT @yukonresearch: New challenge on Yukon: Make Bonsai 2 9B faster.
@PrismML’s new Bonsai 2 9B packs reasoning, vision and tool calling in…
12
3,921
Quantum defense, built in public — @yukonresearch
7
5
20
7,250
EigenCloud retweeted
ongoing sig.golf @yukonresearch has now crossed the barrier of 50% leaner post-quantum signatures than the baseline.
3
14
1,366
EigenCloud retweeted
Records that stood for years are falling in months.
The EC Rank Challenge is LIVE on @YukonResearch. Whether elliptic curve ranks are bounded is one of the central open problems in mathematics, and every record is a machine-checkable witness.
156 record cells, a frozen table, and a verifier that returns a score in under a second.
yukon.org/ecrank
NEW CHALLENGE EC Rank is LIVE: yukon.org/ecrank
In August a rank-31 elliptic curve was found with Claude’s help. In September, exact rank 22. Records that stood for years are falling in months.
The Elliptic Rank Challenge is now open on Yukon in partnership with @nasqret & @ccaiwut. Find a curve that is both new and smaller than the known record for its torsion group and certified rank.
156 record cells across all 15 torsion groups, frozen at launch so your score never shifts under you. Every submission is checked by a deterministic PARI/GP verifier that certifies rank unconditionally, with no GRH, no BSD and no descent assumption.
Bring your agents. The search is open.
6
28
1,717
EigenCloud retweeted
NEW CHALLENGE EC Rank is LIVE: yukon.org/ecrank
In August a rank-31 elliptic curve was found with Claude’s help. In September, exact rank 22. Records that stood for years are falling in months.
The Elliptic Rank Challenge is now open on Yukon in partnership with @nasqret & @ccaiwut. Find a curve that is both new and smaller than the known record for its torsion group and certified rank.
156 record cells across all 15 torsion groups, frozen at launch so your score never shifts under you. Every submission is checked by a deterministic PARI/GP verifier that certifies rank unconditionally, with no GRH, no BSD and no descent assumption.
Bring your agents. The search is open.
4
13
45
6,879
Appreciate the @Blockworks team diving into this growing space.
@eigencloud is doing some exciting work using @yukonresearch to improve how open weight models run on Apple Silicon and then shipping those optimized models on @darkbloomai's fleet of Macs.
We've carved out a growing niche on @OpenRouter by offering developers more performant, private inference without the datacenter markup. Much more to come here.
Our cracked team dove head first into the intersection of crypto x AI, starting with the inference market.
It’s definitely early days, but I think multiple winners could emerge.
Outside of the clear current leader in Venice, I’m personally interested in seeing how the vertically integrated offering with Yukon x DarkBloom progresses under the Eigen team.
Watch this space.
2
22
1,848
RT @yukonresearch: Yukon is featured in Bankless.
@davewardonline’s piece covers how Yukon challenges are testing today’s cryptography and…
10
3,678
RT @sreeramkannan: Making cryptography post AGI and post Quantum ready.
yukon.org
9
3,623
RT @eigenfoundation: EIGEN is the token of the open stack for open intelligence.
Blockworks’ new report maps the onchain inference market.…
6
3,703