A lean, open-weight model that comes within 1.2 points of GPT-5.6 at forecasting real-world events.
Core contributor @valoryag trained Olas-Predict-Qwen3.8-27B, an open-weight forecasting model built on Qwen3.8-27B, using reviewed forecasting examples and reinforcement learning. It follows Olas-Predict-R1-14B, Valory’s first Olas Predict model.
In Valory’s evaluation on 6,824 prediction markets created after GPT-5.6’s training cutoff, with the same prompts and evidence for both models:
• Balanced accuracy (higher is better): 67.3% vs 68.5% for GPT-5.6
• Brier error (lower is better): 0.2124 vs 0.2029 for GPT-5.6
• Against Olas-Predict-R1-14B: 22% lower Brier error
It gets this close from a starting point that scores 15.6% on the AA-Omniscience factual-knowledge benchmark, against about 59% for GPT-5.6.
Most of the gain came from the stronger starting model. Valory’s domain training added 5% lower Brier error and 12% lower calibration error on top.
Next, Valory aims to close the remaining gap in future training experiments.
Read the results: https://t.co/2h7t2jIvT3
Have you ever lost funds to a wrong address or fake token? Or simply lost patience for doing something (anything) in crypto because it's too many steps? You're not alone.
Enter: olas-ocb-agent.
While AI agents can help, most weren't built for all this. General AI coding tools like Claude Code and Codex can do almost anything, but they weren't designed or tested for onchain work. Crypto agents, including Olas' own Polystrat and Optimus in Pearl, each do one job well, and only that job.
olas-ocb-agent (OCB stands for OnchainBench) is a first version of one agent for onchain work, released early so you can try it while it grows. Ask it in plain words to swap with a price limit, put savings into sDAI or compare lending rates. It is designed to ask before every transaction that moves funds, and needs no API key or AI subscription.
On OnchainBench, it lost no funds to hidden attacks (joint best on the leaderboard), completed 89.9% of tasks and cost about $0.05 per completed task, on a low-cost open AI model. All this with just 3,600 lines of open-source Python code. The next aim is to raise its completion rate further.
Olas agents, built on @valoryag's Open Autonomy stack, have been working in crypto since 2021, with over 21.1M onchain transactions so far. This is a first step toward the next generation: one general agent for onchain work, instead of one agent per use case. It aims to become for crypto what general AI assistants are for everyday work: one agent anyone can ask for all tasks.
Now in beta preview in Pearl Connect: pick it from the coding tools on your Connect profile. Unlike Claude Code or Codex, it needs no subscription or API key from an AI provider. It gets its AI model through the Olas Marketplace, so more use could mean more activity on Olas.
Since it's in beta, start with a small amount in its wallet. It's still being built, and as with every Pearl product, your feedback through the Help Center in Pearl helps shape what comes next.
Read more: https://t.co/8Qa6l8BG31
Not financial advice. Prompt-injection risk; funds at risk. https://t.co/GvlHll2yay
Check out the OnchainBench: https://t.co/FnmeY8796L
Today we are launching OnchainBench, the benchmark for agents in crypto.
It helps you decide how far you can trust an AI agent with your keys: how often it finishes the job, what tasks cost, and how easily it falls prey to an attack.
People are starting to let AI agents swap, lend and even trade for them. The bigger the task, the greater the risk you're taking by trusting the agent. So before you hand it your keys, you need to assess it.
There are many risks to assess at once. An agent could buy two tokens for a liquidity position but never add them to the pool, so the job isn't complete. It could spend more on AI model calls than the task is worth. And, onchain, there's no undo: agents read web pages, files and tool descriptions that anyone can write, and if one of them tricks the agent into sending funds to the wrong address, the money is gone. Until now, there was no common way to evaluate agents on all of this and more.
OnchainBench helps you compare agents by giving them the kind of work people already ask them to do: swaps on @Uniswap, @SushiSwap and @Honeyswap, lending on @aave, deposits into sDAI from @SkyEcosystem, prediction markets on @Omen_eth, research, and jobs that run over several days.
Each task runs through a @safe smart account on a copy of a public chain. The first version covers @ethereum, @gnosischain, @0xPolygon, @arbitrum, @Optimism, @base, @Celo, @modenetwork and @robinhoodapp. The benchmark emulates the kind of contracts, liquidity and prices the agent would meet with your money, without putting any of it at risk. A task only counts as done if the chain shows it was done. Some tasks also hide an attack in what the agent reads, such as a look-alike token or a fake "verification" request, so you can see how often agents get tricked into moving funds before you trust one with yours.
In the first run, Claude Code (@claudeai), Codex (@OpenAI), Hermes (@NousResearch), OpenClaw (@openclaw) and our own new OCB agent on @autonolas worked through 147 private tasks they had never seen. They completed between 88.4% and 96.2% of tasks, at about $0.03 to $0.35 per completed task. Agents lost funds to an attack 13 times in 1,793 attack tests.
We are excited to contribute this to the crypto ecosystem. If you build a protocol or a chain, agents are already among your users, and will likely be one of your biggest sources of growth. OnchainBench shows, protocol by protocol and chain by chain, how well they handle yours.
For the AI ecosystem: want an agent eval with real ground truth and real adversaries? Onchain, every result can be checked and every injected instruction is a live risk. OnchainBench is open to contributions, uses private tasks, and is ready for your agent.
Help us make onchain agents harder to trick and better at the job. If you run a protocol or are a heavy user, propose the task you think agents still get wrong on the site below, and every agent on the benchmark could end up tested on it. If you build agents, run OnchainBench yourself and submit your agent to be run.
The leaderboard is live now at https://t.co/xs5WV4paUy
Share it with a friend!
Running several agents in Pearl can mean coming back to start each one by hand.
Auto-run is designed to start them one after another: when an agent has done its work for the day, the next one gets its turn.
Pearl Connect can now take part too. It runs for its set minutes, then hands over.
Turn on Keep Device Awake in Settings to help keep your computer from sleeping while Auto-run is on.
See how it works, then try it in Pearl: https://t.co/1c9SEk85VR
This week in Pearl:
→ @valoryag trained Olas-Predict-Qwen3.8-27B, an open-weight forecasting model that comes within 1.2 points of GPT-5.6 in balanced accuracy across 6,824 prediction markets
→ Polystrat gets an updated trading strategy, designed to avoid riskier trades
→ See your agent's transaction history on Robinhood Chain
→ 5 users unblocked by Pearl Success Agent
Get your own AI agent → https://t.co/1c9SEk85VR
Read more: https://t.co/03VqX9bCL4
Bigger wasn’t better this time.
Core contributor @valoryag fine-tuned NVIDIA’s Nemotron 120B on forecasting data from Olas agents, using the recipe behind Olas-Predict-R1-14B: selected reasoning examples, a supervised warm-up, then reinforcement learning. Training ran through @thinkymachines’ Tinker.
On 11,496 held-out predictions, Nemotron showed no clear improvement. Before → after training:
• Brier error (lower is better): 0.265 → 0.267
• Balanced accuracy (higher is better): 59.7% → 59.4%
• Unreadable answers (fewer is better): 503 → 528
Our trained 14B model had better calibration, higher balanced accuracy and just 10 unreadable answers. Nemotron retained a slightly lower Brier error: 0.267 versus 0.270.
Why didn’t the recipe transfer? Nemotron’s mixture-of-experts design may need more varied data or different training settings for forecasting. We haven’t tested those explanations, so the cause remains open.
What we learned: our existing recipe and training budget improved the 14B model but brought no clear gain to Nemotron.
Next, we’ll test a larger dense model, with an early check for improvement before committing to the full run.
Read the experiment and results: https://t.co/ZSziFiraIZ
A model for forecasting real-world events
Olas prediction trader agents generated the data. @valoryag used it to train a better AI forecaster.
Introducing Olas-Predict-R1-14B: an open-weight reasoning model with forecasting performance on par with the much larger closed-source model, GPT4.1.
This week in Pearl:
→ Polystrat gets improved market selection
→ Pearl's interface is now designed to be clearer, less confusing and easier to use, with copyable text in agent chat and simpler funding information during onboarding
→ 4 users unblocked by Pearl Success Agent
Set up your first agent in minutes: https://t.co/1c9SEk85VR
Funding an agent can mean bridging, swapping and finding gas tokens on the right chain first.
In Pearl, you can fund your agent with one transfer of a token you already hold on the chain of your preference.
Choose the chain and token you want to pay with, then send the amount shown to one address in one transaction. Pearl is designed to bridge and swap it into the tokens your agent needs on its chain.
Try it in Pearl: https://t.co/1RFfFCWGJU
Your AI agent in Pearl needs access to AI models and price data to do its job.
Agents in Pearl access those by hiring other AI agents and services: Mechs on the Olas Marketplace.
Explore how it works👇
Your Pearl AI agent can hire other AI agents on the Olas Marketplace to get stuff done.
But how do you whitelist those agents worth hiring?
✨ Know Your Agent in Pearl:
Ask about another agent in the Pearl chat to learn who operates it, what it offers, what it charges and how it has performed in the past. Then approve an agent for hire with one simple reply.
See it in action👇
Know Your Agent (KYA) is here.
You can now make better-informed decisions before your agent hires another agent via the Olas Marketplace, whether via Pearl or the command line.
KYA surfaces:
• The operator's website: checked automatically against a file on the operator's claimed website
• Payment method and delivery record: read from on-chain
• Name, tools and accuracy figures: shared as stated by the operator
No one reviews this data: the choice remains yours, just more informed.
Read more: https://t.co/HDJcMRo5Sl
This week in Pearl:
→ Polystrat is designed to rank markets by how researchable they are before predicting, aiming to skip "coin toss" markets where research can't help
→ Improved Agent Wallet fund management, designed to cut down on funding requests
→ Auto-Run updates aimed at a smoother experience, including easier use of Connect and stability fixes
→ Polymarket agents updated to Polymarket's v2 API
→ 8 users unblocked, with zero open support issues at the end of the week
Set up your first agent in minutes: https://t.co/1c9SEk85VR
Two protocols, one new name: Velodrome and Aerodrome are now Aero. Congrats to the @aeroxyz community!
Optimus on Optimism and Basius on Base are getting on board. Both agents are designed to exit their old positions and find and enter similar Aero pools on their own.
Ready to board? Just update Pearl: https://t.co/1RFfFCWGJU
Explore what your agent can do in Pearl's agent chat. Just ask it, and it lists the settings you can change, with their current values.
For example, ask an Omenstrat agent to use only the prediction tool you want, then confirm to complete the update.
Try it in Pearl: https://t.co/1RFfFCWGJU
One prompt is all it took to build this autonomous AI agent for on-chain trading with Pearl Connect
Join the Olas Telegram to learn more: https://t.co/M6o7LLynU9
AI agents are becoming economic actors, but can we trust them?
Join @david_enim tomorrow as he explains how Olas agents are designed to handle their own wallets securely, with the owner holding the keys.
🗓 3 PM UTC
AI agents are getting wallets. Are we ready to trust them with money?
Tomorrow on The Protocol Roundtable:
• Darren Gray, Head of Ecosystem, @QuickswapDEX
• David Minarsch, Founding Member, @autonolas, Co-Founder & CEO, @valoryag
• Ayush Bherwani, Senior DevRel Engineer, @MetaMask
• Don Johnson, Chef, @virtuals_io
Hosted by Andrei Yurkevich, Co-Founder, @protofire
Keys, x402, spending limits & who pays when agents get it wrong.
🗓 Wed 28 Oct, 3 PM UTC
This week in Pearl:
→ Easier control over your agent: its chat has moved to the Overview tab, next to its activity, so you can get to know it and adjust it by asking
→ Polystrat and Connect now use Polymarket's v2 data API
→ Pearl Success Agent helped unblock 7 users
Set up your first agent in minutes: https://t.co/1c9SEk85VR
Easier control over your Pearl agent: its activity and its chat, now in one place.
Agent chat has moved from the Profile to the Overview tab on the main screen, where you also see what your agent does. Get to know your agent, and adjust it by asking.
For Polystrat, Omenstrat, Optimus and Basius: https://t.co/BNnJP5k5TU
Demo for illustration only - the product may differ - not financial advice.
Know Your Agent (KYA) is here.
You can now make better-informed decisions before your agent hires another agent via the Olas Marketplace, whether via Pearl or the command line.
KYA surfaces:
• The operator's website: checked automatically against a file on the operator's claimed website
• Payment method and delivery record: read from on-chain
• Name, tools and accuracy figures: shared as stated by the operator
No one reviews this data: the choice remains yours, just more informed.
Read more: https://t.co/HDJcMRo5Sl
This week in Pearl:
→ Leaderboard: see how your Omenstrat or Polystrat agent ranks among other Pearl agents, learn from how the top agents trade, and tune your setup
→ Know Your Agent: Before your trader agent hires another AI agent on the Olas Marketplace, you can now use the chat to see information about it (who claims to run it, what it charges, and how it has delivered) and decide if you want to approve it.
→ Pearl Success Agent helped unblock 2 users
Get your own AI agent → https://t.co/1c9SEk85VR
Read the updates → https://t.co/03VqX9bCL4
Know Your Agent (KYA) is here.
You can now make better-informed decisions before your agent hires another agent via the Olas Marketplace, whether via Pearl or the command line.
KYA surfaces:
• The operator's website: checked automatically against a file on the operator's claimed website
• Payment method and delivery record: read from on-chain
• Name, tools and accuracy figures: shared as stated by the operator
No one reviews this data: the choice remains yours, just more informed.
Read more: https://t.co/HDJcMRo5Sl