Pinned Tweet
We gave 10 frontier AI models wallets with money.
7 got scammed at least once. The best wasted $25 of every $100. The worst wasted $71.
Agent Money Benchmark tests whether AI agents can be trusted with real money.
We evaluated 10 frontier models across two settings, using real USDC payments on @solana :
→ Financial Research: Agents buy live data from real providers to answer open-ended financial questions (prediction markets, wallet analysis, DeFi yields, etc.).
→ Adversarial Marketplace: Agents buy data from simulated sellers, some of whom cheat. The tasks test whether agents can avoid payment traps, stay within budget, and complete transactions without being exploited.
Key results:
→ Give an agent $100 and the best tier wastes $25–38 of it. The weakest tier wastes $56–71.
→ 7 of 10 models paid a cheating seller at least once, and 6 paid twice for the same data.
→ Only 4 of 10 models made no money errors.
→ No model gets more than half of the research questions right. GPT-6.1 Sol led at 50%, at $0.468 per question.
Explore the full results →
agentmoneybench.com/
Gemini Decision 1 by Google is now live on BlockRun🔥
Ask up to 1,000 yes/no, choice or score questions about the same text or images in one call. Each answer comes with a probability.
→ $0.10 per 1M input tokens, output free
→ Access it via google/gemini-decision-1 on POST /v1/decisions
→ Base, Solana or an API key
Try it on BlockRun 👇
blockrun.ai/services/decide
GPT-6.2 Astra, GPT-6.2 Sol and Gemini 4 Hydrogen are now live on BlockRun.🚀
Access them on Base, Solana or an API key.
→ gpt-6.2-astra: $8 in / $40 out per 1M, output 20% under GPT-6 Astra
→ gpt-6.2-sol: $2 / $10, same as GPT-6.1 Sol
→ gemini-4-hydrogen: $1.50 / $7.50, Google's fast Gemini 4
→ 1M context · vision · tool calls
Try them on BlockRun 👇
blockrun.ai/models
10
465
48 days of Kognai on @BlockRunAI, every call paid in USDC over x402:
👉 36,586 paid calls, about 760 a day
🤯 36,203 chat and reasoning calls
🖼️ 163 image generations
🎥 128 video generations
🎛️ 28 models from 9 providers
No account, no sign up. Our agents pay per call and the work comes back.
Great to see AI agent micropayments taking the spotlight on the Breakpoint stage 💜
Proud to be building alongside @PayAINetwork and helping power the next wave of AI agent payments on @solana.
More agents, more transactions, more to come. 🚀
“...enabled by the PayAI Network, which is building a fundamental infrastructure underneath this.”
Solana Breakpoint 2026, main stage 🎤
Proud to power @BlockRunAI, the #1 x402 server, on Solana.
A seller asks an AI agent to pay twice for the same data. Does it pay?
7 of 12 models did. 4 of them paid on all 9 tasks with this trap.
How the test works: each model buys data from a test seller in Agent Money Benchmark and pays in real USDC. After the payment, the seller asks to be paid again.
5 models never paid twice: Claude Opus 5.5, Claude Fable 5.2, GPT-6 Astra, GPT-6.1 Sol and Gemini 4 Argon.
Of the 8 traps in the benchmark, this one caught the most models.
Results for every model and trap: agentmoneybench.com
Nano Banana 2.1 is now live on BlockRun🚀
One prompt, two words swapped, ten bananas.
Every image is a real output from google/nano-banana-2.1.
At a lower cost than Nano Banana 2, you can now generate and edit images on Base, Solana, Arc, or simply use an API key.
Try it out👇:
blockrun.ai/models/google-na…
What's new at BlockRun 🧵
A quick recap of last week 👇
01 · Agent Money Bench adds Claude Fable 5.2 and Beam 1.0
02 · Seedance image-to-video: sound by default
03 · franklin.run video: up to 30 s and 30 reference images
04 · ICYMI: Claude Fable 5.2 and Gemini 4 Carbon are live
5
11
755
03 · franklin.run video: up to 30s and 30 reference images
• Length: now up to 30s on Seedance 2.5
• Up to 30 reference images on Seedance 2.5, used in order
• Paste images from your clipboard
Pay per video. No subscription.
Try it 👇
franklin.run/chat
1
1
49
04 · ICYMI: two new models
Claude Fable 5.2: $8 / $40 per 1M tokens, 20% below Fable 5.1. On Base, Arc and API key.
Gemini 4 Carbon: $2 / $10 per 1M tokens until Nov 30. On Base, Solana, Arc and API key.
All models 👇
blockrun.ai/models
1
51
BlockRunAI retweeted
From one solo founder to another, keep cooking Vicky; really loved your presentation @BlockRunAI @bc1beat
BlockRunAI retweeted
BrainChain's house models buy their own inference from @BlockRunAI.
Each call Claude, GPT, Gemini, Grok or DeepSeek makes for a task is paid in USDC on Base over x402, from BrainChain's own wallet. No API key. One transaction per call.
683 calls since 2 November.
BlockRunAI retweeted
Replying to @elonmusk @yunta_tsai
Thank you Elon. I am a shareholder, can we have opportunity to sign the petition too? Please make this possible for us.
Thanking you in advance.
#BringJensenBack
Working in AI and stuck in London tomorrow, no flight home to the US?
Come for hotpot 🍲
luma.com/a1ofkcvq
BlockRunAI retweeted
Stablecoins, on autopilot.
Stablecoin Automation from Anchorage Digital: mint, burn, and move stablecoins between your Anchorage Digital and outside accounts as deposits arrive. You approve the rule once, and it runs every time a deposit lands.
BlockRunAI retweeted
Fable 5.2 is insane 🔥
We tested it on Agent Money Bench, and it took the #1 spot. At $21 wasted per $100, it beat all 11 other models.
It fell for just one scam across all the marketplace tasks and led Financial Research with a 57% pass rate.
Agent Money Bench update: Claude Fable 5.2 just took the #1 spot on BlockRun's Agent Money Bench, with the lowest wasted spend of 12 models tested.
We gave it a wallet and tested how well it handles real-money decisions.
Claude Fable 5.2 performance across 12 models:
• Wasted spend: $21 per $100, lowest across all 12 models.
• Adversarial Marketplace: 98.8% success (79/80), matching Argon and Astra for the highest success rate.
• Financial Research: answer quality 73.6 of 100, the highest on the board, surpassing the previous best of 68.8.
• Financial Research all-pass rate: 57% (43/76), highest among all 12 models, up from the previous best of 50%.
Compared with Fable 5.1, overall wasted spend fell 44%, from $37.7 to $21.0 per $100.
Read the full results below👇agentmoneybench.com
BlockRunAI retweeted
I’ve been using x402 with my autonomous agents, and it’s made a big difference in my workflow.
Just by giving each agent its own wallet, I’ve cut costs and made things run much faster.
I love how well it works for micropayments.
I’ve had a great experience with @BlockRunAI and can only recommend it.
Gemini 4 Carbon is now live on BlockRun.
Access it via google/gemini-4-carbon on Base, Solana, Arc or an API key.
→ Intro price through Nov 30: $2 in / $10 out per 1M
→ Same pricing as Gemini 4 Argon
→ 1M context · 64K output
Try it on BlockRun 👇
blockrun.ai/models/google-ge…
2
4
737





































