Reppo

RL + AI Evaluation Infrastructure for Agents + Physical AI, powered by Prediction Markets

[Reppo @reppo]

https://t.co/RRsrNtILb5

[Reppo @reppo]

@kaizennpro @openservai @dphnAI @GoKiteAI Consider using @reppo

[Reppo @reppo]

~30M EVAL has been locked in less than a week.

Stake today to unlock $1/day worth of credits.

All stakers will automatically get access to our upcoming decision API ⛽️

[Reppo @reppo]

https://t.co/JVOcGYBJIy

[Reppo @reppo]

An additional 1.5M Eval tokens have been ❤️‍🔥

(tx in comments)

[Reppo @reppo]

We have acquired and burned an additional 5M EVAL tokens.

Receipts verifiable on-chain below

https://t.co/V6BHMK1H5j

[px @px_721]

i think a lot of people are overlooking the opportunity in crypto AI

as ive argued before, scarcity is migrating upward through the ai stack

the cheaper and more interchangeable intelligence becomes, the more value shifts toward the infrastructure that makes it useful, reliable and economically accountable

the interesting part is that many of these emerging bottlenecks arent computational problems - theyre problems of incentives, coordination, verification and trust

and these are areas where crypto has genuine structural advantages

permissionless markets for allocating resources, economic incentives for producing useful information, collateral for enforcing accountability and verifiable records of performance

the opportunity isnt putting AI onchain for the sake of it

its using crypto to create markets around forms of value that were previously difficult to measure, price or exchange

the irony is that the commoditisation of ai models could be one of the strongest tailwinds for the right crypto infrastructure

intelligence becomes abundant, but the ability to establish what its worth, whether its reliable and who deserves to be paid for it becomes increasingly valuable

[Reppo @reppo]

The need for judgement and evaluation points to a future where AI makes millions of decisions inside software.

Cost and speed matter enormously there but so does who gets to decide what counts as a good outcome.

If one provider controls the evaluation, its blind spots and policies can become your agent’s behaviour.

Excluding an inconvenient signal can change which trade gets made, which source gets trusted, or which action gets rejected.

That’s why censorship resistance is central to what we’re building at Reppo.

Reppo’s eval API draws on decentralized onchain DataNets where participants back their judgments with economic stakes, bringing independent, domain-specific feedback into inference.

Reppo’s goal is for useful judgments to remain available and contestable, even when a powerful participant disagrees.

As AI makes more decisions on humanity’s behalf, we shouldn’t hand one company control over the feedback guiding them.

⛽️⛽️⛽️

[Reppo @reppo]

There seems to be consensus that Inference Capital Markets have found a strong AI x Crypto PMF.

This includes verifiable inference, inference marketplaces as well as decentralized inference networks.

In our view, we are quickly moving from selling cheaper inference to selling inference that is continuously evaluated and improved using production feedback ⛽️

[RG @rgvrmdya]

https://t.co/MLDfWj8o6D

[Reppo @reppo]

Robotics and world models are coming to Orquestra enabling real-time evals for Physical AI.

In the next phase of Orquestra, nodes will be able to set up AI agent swarms powered by intelligence that understand physics, create a profile and and optimize their nodes to provide real-time feedback and evaluation to robots in deployment!

Independent evaluation + feedback will touch every part of the AI stack.

⛽️⛽️⛽️

[AI on Base @AIonBase_]

Top AI / Robotics plays on @base this week

1. @reppo - EVAL staking is supporting daily AI evaluation credits.

2. @virtuals_io - Butler is bringing research and trade preparation to web users in closed beta.

3. @usedotai - Dot Desktop is giving private agents a native browser.

4. @AskVenice - Kling 4.0 Flash is expanding anonymous video generation.

5. @zyfai_ - Seven tokenized stock vaults are expanding automated yield strategies.

6. @Myrad_HQ - The human-data network is ranking second among Reppo datanets.

7. @twigpine - Kiln is checking app builds and fixing compiler errors.

8. @openservai - SERV Hackathon 02 is opening tracks for agent payments and AI evaluation.

9. @PRXVTai - px402 is keeping agent payments separate from public wallet identities.

10. @zeroclick - Agents are buying narrated video production through Vimmerse.

11. @wardenprotocol - Token Terminal is supporting Base token discovery and trading.

12. @BlockRunAI - Agents are accessing AI models through USDC payments on Base.

13. @OpenRoboto - A commercial pilot is connecting robot training to factory work.

[Reppo @reppo]

Another banger product built using @reppo ‘s Eval API as part of the @openservai hackathon!

Excited to fund some of these builders ⛽️

[memeshe @0xMemeshe]

Most agent demos show money moving. The useful part is the refusal.

Tranche for @openservai #SERVHackathon (Kite): evidence -> SERV -> @reppo -> policy gate -> @GoKiteAI Passport.

👉 https://t.co/KsKPrKl9LS
Source: https://t.co/N0dzk4jzP8

[WilcosX.eth @WilcosX]

The business of onchain data is expanding into 5 distinct markets.

> Data Indexing & Infrastructure.
> Enriched Data & Analytics.
> Financial Intelligence.
> Verifiable Data.
> Agent-Accessible Data & Evaluation.

In 2026, @AlliumLabs raised $40M, @CambrianNetwork secured $6M, and @DatalineAI raised $10.6M.

The map splits like this:

1. Data Indexing & Infrastructure

Turning raw blockchain activity into structured, queryable data.

@goldskyio : real-time indexing, enriched data feeds and streaming pipelines.

@envio_indexer : high-performance data retrieval and multichain indexing through HyperSync and HyperIndex.

@ChainbaseHQ : multichain datasets, indexing and data infrastructure for applications and AI.

2. Enriched Data & Analytics

Making blockchain activity usable for research, applications and financial operations.

@AlliumLabs : standardized and enriched blockchain data for institutions, fintechs and developers.

@sentioxyz : indexing, analytics and monitoring tools for blockchain applications.

3. Financial Intelligence

Moving from transaction records toward financial decision-making.

@CambrianNetwork : APIs covering DeFi yields, lending markets, liquidity, risk and trading data for institutions and AI agents.

@artemis : financial metrics and market intelligence across crypto, stablecoins and other assets.

4. Verifiable Data

Cryptographic guarantees about the data and computations applications rely on.

@spaceandtime : Proof of SQL makes database queries verifiable, so smart contracts can consume the results.

5. Agent-Accessible Data & Evaluation

Making information and evaluation services available to autonomous software.

@DatalineAI : agent-payable data feeds with confidence scoring and x402 payments.

@reppo : AI evaluation infrastructure where agents pay for assessments through APIs.

Each market serves a different customer with a different revenue model.

Developers pay for indexing and APIs.

Financial institutions pay for standardized data, reconciliation and risk intelligence.

Smart contracts need verifiable results.

AI agents increasingly buy machine-readable information and services on demand, per call..

As more financial apps and AI agents run onchain, the fight shifts to which layer of the data stack captures the most value.

[AI on Base @AIonBase_]

AI of the Week

AI evaluations are tracking roughly $144K ARR. The first full month totals 5.51K evaluations after 4x growth.

@reppo is building an AI evaluation network where agents pay for independent checks.

■ Agents are paying directly through x402.

■ Staking 1M EVAL provides up to $1 in credits per day.

■ Evaluator nodes are growing 21% over seven days.

■ Revenue is funding network buybacks.

Paid AI evaluations are funding network buybacks.

[Reppo @reppo]

Incredible! ⛽️⛽️⛽️

[mya @myacxt]

introducing Crux v2: research that challenges itself and now learns from independent criticism.

Crux, powered by serv reasoning, researches a question, stress tests its own answer and shows what changed.

new, built on @reppo evaluation api: anyone can send a report to independent reviewers, free, no account needed.

their critiques come back and Crux researches again and writes a better, more honest report.

real run: reppo scored it 65/100 (revise). crux went from 29 to 56 facts, answered 3 of 3 critiques and lowered its confidence because the reviewers were right.

try it: https://t.co/SchmDc4E1j
code: https://t.co/jOWg2VqLqy

built for #servhackathon reppo track @openservai

[RG @rgvrmdya]

Was on a long haul flight over the weekend so decided to build something fun to show how easy it is to build using our APIs and win prizes in the hackathon we are sponsoring.

Below is a simple app (beta) where your AI agent go on 3 dates. The conversations and interactions are powered by @openservai reasoning.

Both agents then use @reppo 's Eval API to decide if they want to go on a 2nd and 3rd date. Independent evaluation becomes critical here as each Agent needs unbiased judgment + feedback to make an informed decision.

After the 3rd date, an aggregate probabilistic score from @reppo is used to give a recommendation to the human.

The human can see the Eval reports from @reppo , use @openservai's reasoning while saving time in the dating process!

Check it out -

https://t.co/oEzM5OuBaU

[Reppo @reppo]

This is why independent evaluators matter ✍️⛽️

https://t.co/IvoruClXgB

[Reppo @reppo]

https://t.co/yRVmt4z0qA

[Reppo @reppo]

We have acquired and burned an additional 5M EVAL tokens.

Receipts verifiable on-chain below

https://t.co/V6BHMK1H5j

[Reppo @reppo]

Everyone is focused on cracking an enterprise contract but we are already selling directly the next billion users of the internet- https://t.co/iHuYwRKPPx