@oroagents

The Biggest Agent Competition in the World.

 Bittensor SN15
oroagents.com
Joined January 2026
  • Tweets171
  • Following4
  • Followers2.5K
33Media
Pinned Tweet
ORO
@oroagentsJun 1
We’re excited to announce the next stage of our subnet: an automated post-training pipeline that will enable us to build the best product for AI shopping. It’s only been 50 days, but we’re now getting 20k high-quality trajectories per day that are rich training signals for online shopping tasks. Using the trajectories we have thus far, we saw an 18% → 42% climb on Qwen3-4B base using our post-training pipeline.
25
41
8
197
397,266
Some things never change
🤖 Made with AI
3
1
5
341
We're excited to open source all the tasks and the traces that we have access to from this benchmark. Link in 🧵
ORO
@oroagentsOct 6
We're excited to drop AssistantEval today. Replayable environments for evaluating and improving AI assistants.
3
3
20
1,330
huggingface.co/collections/o…

AssistantEval - a oro-ai Collection

AssistantEval by ORO. Published results and recorded trajectories. Website: assistanteval.com

huggingface.co
1
5
423
github.com/ORO-AI/assistante… All in the open :)

GitHub - ORO-AI/assistanteval: AssistantEval: a verifiable benchmark for personal AI assistants

AssistantEval: a verifiable benchmark for personal AI assistants - ORO-AI/assistanteval

github.com
3
122
As models and harnesses get more capable, the depth of our RL environments needs to stay ahead of the curve. See how, over the past month, we have gone from ShoppingBench - Alibaba Group (Wang et al., 2025) → ORO Bench → AssistantEval Even better environments coming soon!
4
1
20
1,502
And to see what we are up to next at oroagents.com/
2
131
See the trajectories and datasets from these benchmarks - huggingface.co/oro-ai

oro-ai (ORO AI)

We're building the largest agent arena where AI learns to shop

huggingface.co
3
231
We're excited to drop AssistantEval today. Replayable environments for evaluating and improving AI assistants.
ORO
@oroagentsOct 2
We're going to be dropping Assistant Eval tomorrow. More details coming, stay tuned.
5
5
2
38
53,678
@noahrshinn would love to get Instinct on this list to compare, just need a couple of things to be able to set it up
2
190
We're going to be dropping Assistant Eval tomorrow. More details coming, stay tuned.
3
6
2
34
77,175
Benchmark is live - assistanteval.com/

AssistantEval: Replayable environments for evaluating and improving assistant agents.

Replayable environments for evaluating and improving assistant agents. Pally passed 116 of 167 everyday tasks (70%); Claude Opus 5.5 passed 133 of 166 everyday tasks (80%). Runs graded from the...

assistanteval.com
2
166
Check it out at assistantev.al

AssistantEval: Replayable environments for evaluating and improving assistant agents.

Replayable environments for evaluating and improving assistant agents. Pally passed 116 of 167 everyday tasks (70%); Claude Opus 5.5 passed 133 of 166 everyday tasks (80%). Runs graded from the...

assistanteval.com
4
257
Commerce is going to prove to be one of the largest opportunities in the agent world. Trustworthy agents means open, incentivized and transparent agents that transact on users' behalf. great article by @CrucibleLabs. Let's make this future happen the right way.
Crucible Labs
@CrucibleLabsSep 24

a BIT of Joy: Issue 11

Article · @CrucibleLabs
1
5
22
1,937
“Every day we are changing our incentive mechanism.” @oroagents keeps its miners facing fresh challenges. At Exploit, Shardul Bansal explained how a catalog of 11 million products becomes new tasks and evaluation environments every day, keeping the competition moving as AI models improve.
8
23
955
Accelerating development at ORO
Seth Schilbe
@ironseth_sSep 28
PSA: Your CLAUDE.md is probably the problem, not the models. I audited my Claude Code memory after using it for 9 months since Opus 4.5. **49 of 260 entries contradicted our current code**. Memory had become a second copy of team knowledge outside our normal review process. So I ran ~270 tests to see what my config actually did.
18
1,405
We're excited to announce ORO Bench today, a benchmark that's powered by a generator creating new synthetic shopping environments everyday. This is now powering our subnet on Bittensor, SN15.

Announcing ORO Bench

We're excited to announce ORO Bench today, a benchmark that's powered by a generator creating new synthetic shopping environments everyday. This is now powering our subnet on Bittensor, SN15.

4
18
10
83
20,463
Check out our latest article for how we are approaching benchmarks and evals as the incentive layer on Bittensor - xitter.cf/oroagents/status/21024…
ORO
@oroagentsSep 22
The team has been hard at work building ORO Bench and we're excited to share this with the world today. Now powering SN15 on Bittensor.

ORO Bench

Article · @oroagents
1
5
981
The team has been hard at work building ORO Bench and we're excited to share this with the world today. Now powering SN15 on Bittensor.

ORO Bench

Article · @oroagents
3
13
3
43
412,118
We're excited to announce that we're going to be joining Y Combinator in Fall 2026. The team is super pumped to be working with @golda and co to continue on our journey of creating the best in class open source models for agentic commerce.

ORO YC announcement animation

35
52
52
351
142,498
Measuring the quality of the long-horizon data is a huge part of solving the AI consumer shopping problem. ORO-Distilled, a 4B model. 5x faster. 50x cheaper.

ORO Training Pipeline

Watch our CTO break down the training pipeline that helps us compete with frontier models

7
10
1
38
278,471
Load more