Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introducing NDI 1.0 - a small model for document processing. #1 on Parse Index Normalized (across 6 public benchmarks) GPT-6 Astra performance at 1/10th the cost 📈 Trained on 15M+ financial files for industry-grade accuracy. Get started with $25 in free credits: Reply "NDI" for a single file setup... show more

Nace AI

1,097 subscribers

1,304,753 views • 9 days ago •via X (Twitter)

Finance Science & Technology

Anya Rossi• Live Now

Private livecam show

43 Comments

Nace AI9 days ago

OmniDocBench

Nace AI9 days ago

ParseBench

Nace AI9 days ago

MDPBench

Nace AI9 days ago

ExtractBench

Kaitee9 days ago

curious how it handles multi-page contracts with tables that span pages. most of these tools still choke on that.

Nace AI9 days ago

Developer Access:

Nace AI9 days ago

Developer Console Access

Veles9 days ago

Already running drex in night desk, NDI for filings is the obvious add

kai.9 days ago

1/10th the cost for financial doc parsing is massive.. specialized small models keep winning

carpe9 days ago

how accurate is NDI on messy real invoices?

Ruuj9 days ago

Making it cost efficient with such great results is a good one guys

Nawi9 days ago

The "every value links back to its source" part is the real flex. An agent you can audit beats an agent you have to trust.

ben.9 days ago

bangerrr!!

marcus9 days ago

love that you shipped a single file setup for Claude Code and Codex. how long does it take to get parsing from zero?

Nace AI9 days ago

Pretty much right away, add your API key, and you can start parsing in a few minutes.

YasuoXBT🥷9 days ago

Ndi

Nace AI9 days ago

check your DMs!

Arman Suleimenov9 days ago

NDI

Nace AI9 days ago

Check your DMs!

Kaif9 days ago

"astra performance at 1/10th the price" is the only benchmark founders actually read

lis9 days ago

trying this

LX™9 days ago

Parse, split, classify, extract, ground. One document goes through all five and stays linked to its source.

CesarAG9 days ago

I'll try it, it will help me a lot in my work

zen.9 days ago

curious how much the 15M+ financial files bias it toward finance. does accuracy hold up on contracts or medical docs?

Abdirakhman Ismail9 days ago

NDI

Nace AI9 days ago

Your DMs are closed, DM us and we’ll send it over!

Alex Ai | Tools & Updates9 days ago

GPT-6 level accuracy at tenth cost on financial docs impressive

banzaan9 days ago

@pouyanroohii

Mour₳‎9 days ago

Ndi

Nace AI9 days ago

Your DMs are closed, DM us and we’ll send it over!

Mira Synth Tech9 days ago

Hitting number one on the Parse Index with that kind of cost efficiency is seriously impressive

Clara AI9 days ago

Affordable, high-accuracy document processing could be a game changer for AI workflows

Ace9 days ago

Everyone's building smarter agents, but half of them are still squinting at blurry PDFs. Fixing the eyes first makes a lot of sense.

Lau R🍕9 days ago

the mountain in that image is doing more marketing than the entire product page

NOOR TECH8 days ago

Outstanding Nace AI launch, NDI 1.0 delivers top accuracy at a fraction of cost efficiently!

LastResort9 days ago

Cost per page? @grok

AM9 days ago

NDI

Nace AI9 days ago

Check your DMs 🙌

Zivlen 지능9 days ago

interested

An Nayal8 days ago

1/10th the cost.. you're cracked 🫡

Ye Cooper9 days ago

NDI

Nace AI9 days ago

Check your DMs!

Aaliya9 days ago

15M files is a lot of data to train on

Related Videos

Introducing Databricks Document Intelligence: a research-specialized layer that turns raw enterprise documents into structured data your agents can actually reason over. Across our benchmarks, Document Intelligence delivered the highest end-to-end parsing and extraction quality at 6-8x lower cost, with a 16% average performance gain across every agent framework tested, just from better parsing. How to get started:

Introducing Databricks Document Intelligence: a research-specialized layer that turns raw enterprise documents into structured data your agents can actually reason over. Across our benchmarks, Document Intelligence delivered the highest end-to-end parsing and extraction quality at 6-8x lower cost, with a 16% average performance gain across every agent framework tested, just from better parsing. How to get started:

Databricks

19,428 views • 6 months ago

holy f**k, this is insane GPT-6 Astra and Claude Fable 5.1 are free on Xplabs right now what you get: - GPT-6 Astra free - Claude Fable 5.1 free - and 1000+ models open on the platform how to try (under 1 min): step 1: go to step 2: search for gpt-6-astra or fable-5.1 step 3: select it and start chatting no plan required, no per-model payment, just pick and go

holy f**k, this is insane GPT-6 Astra and Claude Fable 5.1 are free on Xplabs right now what you get: - GPT-6 Astra free - Claude Fable 5.1 free - and 1000+ models open on the platform how to try (under 1 min): step 1: go to step 2: search for gpt-6-astra or fable-5.1 step 3: select it and start chatting no plan required, no per-model payment, just pick and go

Sakata

45,089 views • 1 month ago

Imagine processing 1,000 medical insurance claims. Each one contains hospital bills, patient records, discharge summaries, and supporting documents. That's a massive amount of unstructured information. Nace Document Intelligence is built to handle the perception part of this workflow. → Parse medical documents → Classify different file types → Extract relevant information → Ground values in their original sources Then Drex can handle the next decision-making step. This is a practical example of why AI agents need more than just powerful LLMs. One CLI. One SDK. Compatibility with Claude Code, Codex, Pi, OpenCode, Hermes, and Prime Agent.

Imagine processing 1,000 medical insurance claims. Each one contains hospital bills, patient records, discharge summaries, and supporting documents. That's a massive amount of unstructured information. Nace Document Intelligence is built to handle the perception part of this workflow. → Parse medical documents → Classify different file types → Extract relevant information → Ground values in their original sources Then Drex can handle the next decision-making step. This is a practical example of why AI agents need more than just powerful LLMs. One CLI. One SDK. Compatibility with Claude Code, Codex, Pi, OpenCode, Hermes, and Prime Agent.

Markandey Sharma

127,970 views • 9 days ago

CTRL+F is not enough anymore. Imagine asking a question across hundreds of PDFs, financial reports, spreadsheets, and contracts. And getting an answer backed by the original documents. That's the idea behind combining three AI layers: NDI → Understands the documents. Drex → Makes decisions using extracted information. GPT → Reasons over the results and explains them. NDI supports 36 file types and is tied for #1 on the parsing leaderboard. Search is evolving from finding keywords to understanding information.

CTRL+F is not enough anymore. Imagine asking a question across hundreds of PDFs, financial reports, spreadsheets, and contracts. And getting an answer backed by the original documents. That's the idea behind combining three AI layers: NDI → Understands the documents. Drex → Makes decisions using extracted information. GPT → Reasons over the results and explains them. NDI supports 36 file types and is tied for #1 on the parsing leaderboard. Search is evolving from finding keywords to understanding information.

JAYDEN™

111,048 views • 9 days ago

Coding agents that actually get better the more your team uses them. Introducing Hivemind: continual learning for AI coding agents. Hivemind turns the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, OpenClaw, Pi) into reusable skills, then pushes those skills across all of them. All on your cloud storage. Now with SkillOpt built in, your skills get trained: +19.1 points of accuracy in Claude Code, +24.8 in Codex, best or tied on all 52 setups tested. Open source, one line install.

Coding agents that actually get better the more your team uses them. Introducing Hivemind: continual learning for AI coding agents. Hivemind turns the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, OpenClaw, Pi) into reusable skills, then pushes those skills across all of them. All on your cloud storage. Now with SkillOpt built in, your skills get trained: +19.1 points of accuracy in Claude Code, +24.8 in Codex, best or tied on all 52 setups tested. Open source, one line install.

Davit

423,450 views • 4 months ago

today, we're launching Light Parse Extend can now parse documents up to 70% cheaper than other solutions - light parse starts at $0.00625 per page, and scales down significantly with volume discounts - it's SOTA on Databricks OfficeQA Pro (a 3rd party agent benchmark) among all non-agentic parsing providers our ML/AI team trained smaller and more efficient layout, table, and form vision models to enable significant cost reduction without impacting accuracy on many types of documents live today for all users!

today, we're launching Light Parse Extend can now parse documents up to 70% cheaper than other solutions - light parse starts at $0.00625 per page, and scales down significantly with volume discounts - it's SOTA on Databricks OfficeQA Pro (a 3rd party agent benchmark) among all non-agentic parsing providers our ML/AI team trained smaller and more efficient layout, table, and form vision models to enable significant cost reduction without impacting accuracy on many types of documents live today for all users!

Kushal Byatnal

49,758 views • 1 month ago

I ran the same task on Claude Code and DeepSeek's new agent harness. One cost $150. The other cost $2. Today we're launching (AgentSky), the "OpenRouter for Agents" — one API → Claude Code, Codex, DeepSeek, Kimi, OpenCode, and every major agent in the cloud. And Agent Playground on top: race them on your own task, with your real tools (GitHub, Gmail, more), side by side in a browser: time, cost, tokens burnt. Guess which one was $2.

I ran the same task on Claude Code and DeepSeek's new agent harness. One cost $150. The other cost $2. Today we're launching (AgentSky), the "OpenRouter for Agents" — one API → Claude Code, Codex, DeepSeek, Kimi, OpenCode, and every major agent in the cloud. And Agent Playground on top: race them on your own task, with your real tools (GitHub, Gmail, more), side by side in a browser: time, cost, tokens burnt. Guess which one was $2.

Xiaoyin Qu

2,226,054 views • 1 month ago

introducing operator-1, an autonomous document extraction agent operator-1 saturates multiple 3rd party benchmarks, and delivers the most accurate performance on complex tasks that other solutions can't handle: > long tables with thousands of rows > lengthy files with context scattered across hundreds of pages > dense diagrams and images > much more! operator-1 is a specialized agent we built that: - runs within a custom workspace - knows how to use dedicated tools (parse, search, zoom, code execution, validation) to handle any document - learns via persistent memory

introducing operator-1, an autonomous document extraction agent operator-1 saturates multiple 3rd party benchmarks, and delivers the most accurate performance on complex tasks that other solutions can't handle: > long tables with thousands of rows > lengthy files with context scattered across hundreds of pages > dense diagrams and images > much more! operator-1 is a specialized agent we built that: - runs within a custom workspace - knows how to use dedicated tools (parse, search, zoom, code execution, validation) to handle any document - learns via persistent memory

Kushal Byatnal

52,601 views • 10 days ago

Official OpenAI tip for Codex: stop burning GPT-6.1 Sol context on tasks GPT-6 Luna can swarm, while GPT-6 Astra sits idle put Astra on call as a dedicated adversary codex --model gpt-6.1-sol --agent adversary GPT-6.1 Sol acts as the lead architect, scoping architecture and validating pull requests GPT-6 Luna teammates take isolated git worktrees and hammer out implementations at medium effort GPT-6 Astra monitors the peer-to-peer bus as a dedicated adversary, only stepping in at three critical moments: → before a contract locks: does the frontend payload match the backend schema? → when a test breaks twice: are we patching the bug or just hiding the symptom? → before calling done: what attack surface or edge case did everyone overlook? GPT-6 Luna builds. GPT-6 Astra challenges. GPT-6.1 Sol merges and ships Jev engineering runs the identical pattern one layer down: mechanical decisions that need no reasoning (which file to open, which tool to invoke, retry or abort) execute in 16ms, so the frontier models only wake up when execution paths actually diverge Plan on high. Delegate on medium. Keep the adversary on peer-to-peer call. - the full team setup > GPT-6.1 Sol on high coordinates the session and synthesizes PRs > teammate-ux builds client state inside a separate git worktree > teammate-back writes endpoints and handles Redis atomic locks > teammate-adv runs invariant fuzzing and catches replay vulnerabilities > all teammates communicate directly over peer-to-peer messaging without coordinator drag > tasks.json manages file-level mutex locks so nobody steps on another worktree paste the setup and this prompt into Codex CLI below: "Configure my Codex workspace for autonomous agent teams: 1. Audit ~/.codex/agents and .codex/agents for teammate profiles fitting ux, backend, and adversary > Generate definitions only for missing roles > Pin model: gpt-6-luna, model_reasoning_effort: medium for workers, and model: gpt-6-astra for adversary > Assign each worker an isolated git worktree branch to eliminate file conflicts 2. Enable agent teams in ~/.codex/config.toml: > Set [agents] max_concurrent_threads_per_session = 8 > Lock main session model to gpt-6.1-sol with model_reasoning_effort = 'high' > Set approvals_reviewer = 'auto_review' 3. Initialize tasks.json in repo root with atomic file locking, allowing teammates to claim subtasks and resolve contracts directly over inter-agent messaging without coordinator intervention 4. Add a team rule to AGENTS.md: teammates must request an adversary security review before opening a pull request for the lead session Show every configuration diff first. Do not apply edits until confirmed" ↳

Official OpenAI tip for Codex: stop burning GPT-6.1 Sol context on tasks GPT-6 Luna can swarm, while GPT-6 Astra sits idle put Astra on call as a dedicated adversary codex --model gpt-6.1-sol --agent adversary GPT-6.1 Sol acts as the lead architect, scoping architecture and validating pull requests GPT-6 Luna teammates take isolated git worktrees and hammer out implementations at medium effort GPT-6 Astra monitors the peer-to-peer bus as a dedicated adversary, only stepping in at three critical moments: → before a contract locks: does the frontend payload match the backend schema? → when a test breaks twice: are we patching the bug or just hiding the symptom? → before calling done: what attack surface or edge case did everyone overlook? GPT-6 Luna builds. GPT-6 Astra challenges. GPT-6.1 Sol merges and ships Jev engineering runs the identical pattern one layer down: mechanical decisions that need no reasoning (which file to open, which tool to invoke, retry or abort) execute in 16ms, so the frontier models only wake up when execution paths actually diverge Plan on high. Delegate on medium. Keep the adversary on peer-to-peer call. - the full team setup > GPT-6.1 Sol on high coordinates the session and synthesizes PRs > teammate-ux builds client state inside a separate git worktree > teammate-back writes endpoints and handles Redis atomic locks > teammate-adv runs invariant fuzzing and catches replay vulnerabilities > all teammates communicate directly over peer-to-peer messaging without coordinator drag > tasks.json manages file-level mutex locks so nobody steps on another worktree paste the setup and this prompt into Codex CLI below: "Configure my Codex workspace for autonomous agent teams: 1. Audit ~/.codex/agents and .codex/agents for teammate profiles fitting ux, backend, and adversary > Generate definitions only for missing roles > Pin model: gpt-6-luna, model_reasoning_effort: medium for workers, and model: gpt-6-astra for adversary > Assign each worker an isolated git worktree branch to eliminate file conflicts 2. Enable agent teams in ~/.codex/config.toml: > Set [agents] max_concurrent_threads_per_session = 8 > Lock main session model to gpt-6.1-sol with model_reasoning_effort = 'high' > Set approvals_reviewer = 'auto_review' 3. Initialize tasks.json in repo root with atomic file locking, allowing teammates to claim subtasks and resolve contracts directly over inter-agent messaging without coordinator intervention 4. Add a team rule to AGENTS.md: teammates must request an adversary security review before opening a pull request for the lead session Show every configuration diff first. Do not apply edits until confirmed" ↳

mirku

31,072 views • 15 days ago

evo v0.3 is now out! evo now has RLMs, forked context agents and much more ! read the thread for more details for those unaware : evo helps you run autoresearch on your codebase. its a simple claude code / codex / openclaw / hermes agent plugin that's super easy to install and get started with here's whats new in v0.3 -

evo v0.3 is now out! evo now has RLMs, forked context agents and much more ! read the thread for more details for those unaware : evo helps you run autoresearch on your codebase. its a simple claude code / codex / openclaw / hermes agent plugin that's super easy to install and get started with here's whats new in v0.3 -

Alok Bishoyi

20,696 views • 5 months ago

Astra is awesome, but if you're using it specifically to parse and extract a massive volume of documents, you will break the bank (10c+ per page on parsing and extraction!) We've added a native LlamaParse connector to ChatGPT that gives the model access to high-quality document OCR and extraction at the pareto frontier of accuracy and cost. It contains native access to our parse and extract endpoints. Upload your most complex documents, and you can do the following: * digitalize it into extremely accurate context for any QA task * perform large-scale extraction from a set of documents with accurate grounding and citations Offload the higher-level reasoning to Astra, leave specialized doc parsing and extraction to us. ChatGPT plugin: Come sign up here:

Astra is awesome, but if you're using it specifically to parse and extract a massive volume of documents, you will break the bank (10c+ per page on parsing and extraction!) We've added a native LlamaParse connector to ChatGPT that gives the model access to high-quality document OCR and extraction at the pareto frontier of accuracy and cost. It contains native access to our parse and extract endpoints. Upload your most complex documents, and you can do the following: * digitalize it into extremely accurate context for any QA task * perform large-scale extraction from a set of documents with accurate grounding and citations Offload the higher-level reasoning to Astra, leave specialized doc parsing and extraction to us. ChatGPT plugin: Come sign up here:

Jerry Liu

11,083 views • 1 month ago

We just launched Agentsky AgentSky, world’s 1st Agent Market! OpenRouter is for models. AgentSky is for agents. Use 40+ agents—Claude Code, Codex, OpenCode, Hermes, Pi in your browser(even your phone!) or via one API. All without installing or setting up anything. Hit Codex Astra’s weekly limit? Hand off to another agent such as OpenCode + DeepSeek V4.1 in browser without losing any context. You can compare any agent + model directly in browser and that's how I found Astra costs $5.3 while deepseek v4.1 cost $0.12 on the same dashboard task. (I actually preferred deepseek) Try it at If you can try any agent easily without install, which agent do you want to try next? Type the agent name and I will give $20 credit (you can use it for any model)

We just launched Agentsky AgentSky, world’s 1st Agent Market! OpenRouter is for models. AgentSky is for agents. Use 40+ agents—Claude Code, Codex, OpenCode, Hermes, Pi in your browser(even your phone!) or via one API. All without installing or setting up anything. Hit Codex Astra’s weekly limit? Hand off to another agent such as OpenCode + DeepSeek V4.1 in browser without losing any context. You can compare any agent + model directly in browser and that's how I found Astra costs $5.3 while deepseek v4.1 cost $0.12 on the same dashboard task. (I actually preferred deepseek) Try it at If you can try any agent easily without install, which agent do you want to try next? Type the agent name and I will give $20 credit (you can use it for any model)

Xiaoyin Qu

2,725,930 views • 29 days ago

i found a website that gives you the best AI models for free, every single day 😳 claude, gpt, gemini, deepseek and 10+ more, all in one place. no subscription, no credit card. it's called poe. you get free credits every day, and they reset every 24 hours so you never really run out. what you can use for free: - claude, for writing and thinking through hard stuff - gpt, for everyday questions and chat - gemini, for long documents and images - deepseek, for tricky problems - and 10+ more, all in one login how to start (2 minutes): > 1. go to (or get the app on your phone or mac) > 2. sign up with google or email, no card needed > 3. you land straight in the chat, free credits already there > 4. pick a model on the left and start typing > 5. switch models anytime, the credits refill every day one simple tip to make your credits last: use the small fast models for easy stuff, and save the big ones for when you really need them. don't waste a powerful model on a yes or no question. what this saves you: - chatgpt plus: $20/mo - claude pro: $20/mo - gemini advanced: $20/mo that's $60 a month for three apps that mostly do the same thing. poe gives you all of them in one place, for free. one login. every top model. zero dollars. bookmark this so you stop paying for AI you can get for free

i found a website that gives you the best AI models for free, every single day 😳 claude, gpt, gemini, deepseek and 10+ more, all in one place. no subscription, no credit card. it's called poe. you get free credits every day, and they reset every 24 hours so you never really run out. what you can use for free: - claude, for writing and thinking through hard stuff - gpt, for everyday questions and chat - gemini, for long documents and images - deepseek, for tricky problems - and 10+ more, all in one login how to start (2 minutes): > 1. go to (or get the app on your phone or mac) > 2. sign up with google or email, no card needed > 3. you land straight in the chat, free credits already there > 4. pick a model on the left and start typing > 5. switch models anytime, the credits refill every day one simple tip to make your credits last: use the small fast models for easy stuff, and save the big ones for when you really need them. don't waste a powerful model on a yes or no question. what this saves you: - chatgpt plus: $20/mo - claude pro: $20/mo - gemini advanced: $20/mo that's $60 a month for three apps that mostly do the same thing. poe gives you all of them in one place, for free. one login. every top model. zero dollars. bookmark this so you stop paying for AI you can get for free

Isra

58,018 views • 3 months ago

Congratulations to the Cohere team on releasing North Mini Code 1.0 🔥 We provide day-0 support on MLX thanks to our close collaboration with the Cohere team. North Mini Code 1.0 is a 30B param MoE model with 3B active. It runs at ~66 tok/s in BF16, truly impressive speeds before any compression. Here is a quick demo I built using Pi by Mario Zechner & Armin Ronacher ⇌ as my agent hardness. Get started today: 1. Install mlx-vlm from source: > uv pip install git+ 2. Start the server and power your favourite agents (pi, opencode, hermes and etc) > uv run mlx_vlm.server --model CohereLabs/North-Mini-Code-1.0 New release tomorrow

Congratulations to the Cohere team on releasing North Mini Code 1.0 🔥 We provide day-0 support on MLX thanks to our close collaboration with the Cohere team. North Mini Code 1.0 is a 30B param MoE model with 3B active. It runs at ~66 tok/s in BF16, truly impressive speeds before any compression. Here is a quick demo I built using Pi by Mario Zechner & Armin Ronacher ⇌ as my agent hardness. Get started today: 1. Install mlx-vlm from source: > uv pip install git+ 2. Start the server and power your favourite agents (pi, opencode, hermes and etc) > uv run mlx_vlm.server --model CohereLabs/North-Mini-Code-1.0 New release tomorrow

Prince Canuma

35,367 views • 4 months ago

Over 1 billion PDFs are created every day, but your agents still can’t read them reliably. Today we’re releasing Parse 2.0, the most accurate document parsing API in the world. Extend already processes millions of pages daily for leading AI teams like Brex, Mercury, Opendoor, Flatiron Health, and hundreds of others. Now, its even better. Parse 2.0 is SOTA quality on RealDoc-Bench, our open source benchmark that measures agent success rate on real world docs that agents actually encounter in production. We trained Parse 2.0 on 1M+ pages of the hardest documents seen in production. Here’s how it stacks up: - #1 in healthcare, real estate, logistics, and financial services - 95.7% agent Q&A accuracy on 581 docs (next best: 92%) - 0.847 F1 on layout (next best: 0.759) Give it a try today and build production-ready document agents with Extend.

Over 1 billion PDFs are created every day, but your agents still can’t read them reliably. Today we’re releasing Parse 2.0, the most accurate document parsing API in the world. Extend already processes millions of pages daily for leading AI teams like Brex, Mercury, Opendoor, Flatiron Health, and hundreds of others. Now, its even better. Parse 2.0 is SOTA quality on RealDoc-Bench, our open source benchmark that measures agent success rate on real world docs that agents actually encounter in production. We trained Parse 2.0 on 1M+ pages of the hardest documents seen in production. Here’s how it stacks up: - #1 in healthcare, real estate, logistics, and financial services - 95.7% agent Q&A accuracy on 581 docs (next best: 92%) - 0.847 F1 on layout (next best: 0.759) Give it a try today and build production-ready document agents with Extend.

Kushal Byatnal

589,392 views • 4 months ago

Baidu Inc. just dropped PaddleOCR-VL-1.5, and it’s not just another OCR model. This is a fully open-source OCR model with just 0.9B parameters, built specifically for production-grade document understanding. And the results speak loudly 👇 On OmniDocBench V1.5, the most authoritative global document parsing benchmark, PaddleOCR-VL-1.5 ranks #1 overall, hitting 94.5% overall accuracy, outperforming models like DeepSeek-OCR2. As the first OCR model natively supporting irregular document layout positioning, it perfectly solves OCR pain points in production for complex structured business documents such as financial reports and insurance forms. While a lot of the industry is racing toward bigger models, PaddleOCR is betting on something smarter: small parameters + high performance + open source + real usability. That combination is exactly what scalable document intelligence needs right now. Demo / API → Open source → Model → Curious to see how teams start building on top of this.

Baidu Inc. just dropped PaddleOCR-VL-1.5, and it’s not just another OCR model. This is a fully open-source OCR model with just 0.9B parameters, built specifically for production-grade document understanding. And the results speak loudly 👇 On OmniDocBench V1.5, the most authoritative global document parsing benchmark, PaddleOCR-VL-1.5 ranks #1 overall, hitting 94.5% overall accuracy, outperforming models like DeepSeek-OCR2. As the first OCR model natively supporting irregular document layout positioning, it perfectly solves OCR pain points in production for complex structured business documents such as financial reports and insurance forms. While a lot of the industry is racing toward bigger models, PaddleOCR is betting on something smarter: small parameters + high performance + open source + real usability. That combination is exactly what scalable document intelligence needs right now. Demo / API → Open source → Model → Curious to see how teams start building on top of this.

Bishal Nandi

35,318 views • 8 months ago

Claude Code vs Codex vs Pi: which coding agent wins? Melissa Pan, PhD candidate at UC Berkeley’s Sky Computing Lab and previous Arena.ai intern, explored the hidden “harness tax”: how the system surrounding an AI model affects its cost and performance. She reports three surprising findings. One: harness choice impacts cost more than accuracy. To learn about all three findings, and what they mean for building useful coding agents on realistic budgets, check out the full video here:

Claude Code vs Codex vs Pi: which coding agent wins? Melissa Pan, PhD candidate at UC Berkeley’s Sky Computing Lab and previous Arena.ai intern, explored the hidden “harness tax”: how the system surrounding an AI model affects its cost and performance. She reports three surprising findings. One: harness choice impacts cost more than accuracy. To learn about all three findings, and what they mean for building useful coding agents on realistic budgets, check out the full video here:

Arena.ai

85,347 views • 19 days ago

HERMES AGENT NOW RUNS A FULL BUSINESS FOR $8/MONTH. CONTENT. CODE. INBOX. ALL AGENTS. EVERYTHING WHILE YOU SLEEP. one user built this in one session. zero terminal commands. browser only. the team: → content agent: finds video topics, writes first drafts, helps with animations → code builder: writes website and SaaS code → ops agent: organizes and answers email total setup cost: $1. monthly runtime: $8 VPS + cheap model API. the trick is matching model to task. frontier models (Opus 4.8, GPT-5.5) for complex /goals only. $5/million tokens. mid-tier (Claude Sonnet, Haiku, Kimiko 2.6) for daily content and research work. budget (DeepSeek V4 Flash, MiniMax 2.7, Gemma 4, Qwen 3) for cron jobs and routine. run the initial setup on Opus. switch to DeepSeek for everything after. your bill drops 90% with zero quality loss on routine work. the agent team picks up tasks from kanban. each completed task creates a skill. week 3 the same workflows run faster than week 1. full breakdown on Hermes Os in the article 👇

HERMES AGENT NOW RUNS A FULL BUSINESS FOR $8/MONTH. CONTENT. CODE. INBOX. ALL AGENTS. EVERYTHING WHILE YOU SLEEP. one user built this in one session. zero terminal commands. browser only. the team: → content agent: finds video topics, writes first drafts, helps with animations → code builder: writes website and SaaS code → ops agent: organizes and answers email total setup cost: $1. monthly runtime: $8 VPS + cheap model API. the trick is matching model to task. frontier models (Opus 4.8, GPT-5.5) for complex /goals only. $5/million tokens. mid-tier (Claude Sonnet, Haiku, Kimiko 2.6) for daily content and research work. budget (DeepSeek V4 Flash, MiniMax 2.7, Gemma 4, Qwen 3) for cron jobs and routine. run the initial setup on Opus. switch to DeepSeek for everything after. your bill drops 90% with zero quality loss on routine work. the agent team picks up tasks from kanban. each completed task creates a skill. week 3 the same workflows run faster than week 1. full breakdown on Hermes Os in the article 👇

YanXbt

30,910 views • 4 months ago

FREE GPT 5.5 API from a Chinese provider. I already farmed $520 in credits. Plug it directly into Hermes Agent or OpenClaw. > This is a Chinese proxy for OpenAI's official API - same models, same token pricing, same caching. GPT 5.5, Codex, the full stack. I found a way to farm credits through a Telegram bot and it still works. Here's the exact flow: → Go to on Telegram → subscribe to the channel → Join and complete the captcha (use the app, tap the first button) → Go back to the bot, grab your referral link → Invite 1 friend → you get an invite code → Register at with the invite code → Farm points: daily spins (+1-10 pts), each referral (+1/4 pt) → Redeem points for credits: $10 API = 1pt, $100 API = 5pts, $1000 API = 50pts → Create API key → select Codex/another GPT model → plug into Hermes Agent or OpenClaw Ready-made configs for Claude Code and OpenCode are in the "Use Key" section under API Keys. > I have $520 balance. Prices match official OpenAI rates. Caching works. Everything tested. // You can also use these points to buy a ChatGPT account or a subscription for your account. But they get blocked pretty quickly. USE THE API. // Nobody knows how long this lasts. Coupons have been restocked twice already. Move now.

FREE GPT 5.5 API from a Chinese provider. I already farmed $520 in credits. Plug it directly into Hermes Agent or OpenClaw. > This is a Chinese proxy for OpenAI's official API - same models, same token pricing, same caching. GPT 5.5, Codex, the full stack. I found a way to farm credits through a Telegram bot and it still works. Here's the exact flow: → Go to on Telegram → subscribe to the channel → Join and complete the captcha (use the app, tap the first button) → Go back to the bot, grab your referral link → Invite 1 friend → you get an invite code → Register at with the invite code → Farm points: daily spins (+1-10 pts), each referral (+1/4 pt) → Redeem points for credits: $10 API = 1pt, $100 API = 5pts, $1000 API = 50pts → Create API key → select Codex/another GPT model → plug into Hermes Agent or OpenClaw Ready-made configs for Claude Code and OpenCode are in the "Use Key" section under API Keys. > I have $520 balance. Prices match official OpenAI rates. Caching works. Everything tested. // You can also use these points to buy a ChatGPT account or a subscription for your account. But they get blocked pretty quickly. USE THE API. // Nobody knows how long this lasts. Coupons have been restocked twice already. Move now.

Atenov int.

65,251 views • 4 months ago

$$50 free credits for GLM 5.2, KIMI 2.7, and DeepSeek V4 Pro. No promo most people know about. Your buddy pays $20/mo for Claude Pro and burns through it in a week. > You get $50 in credits on Aiand for models that actually beat Sonnet on code benchmarks - and the balance barely moves. Inside: → GLM 5.2 - the model that shipped with better coding scores than closed frontier → KIMI K2.7 Code - long-context beast, sits at the top of open leaderboards → DeepSeek V4 Pro - reasoning at a fraction of GPT-5.6 cost → OpenAI-compatible endpoint - plug into any agent, IDE, or tool you already run Setup, 5 steps: → aiand(.)com → Sign up with Google or GitHub → If the page won't load, switch to an EU VPN and retry → Top right corner - click Credits, select $50 → In Stripe checkout, apply promo code: UNITABETAI → $50 lands. Start firing prompts. Balance drains slowly. I've been running agents on it for days and barely dented it. The promo code is valid as of the time of publication Bookmark this before they cap the promo.$

$50 free credits for GLM 5.2, KIMI 2.7, and DeepSeek V4 Pro. No promo most people know about. Your buddy pays $20/mo for Claude Pro and burns through it in a week. > You get $50 in credits on Aiand for models that actually beat Sonnet on code benchmarks - and the balance barely moves. Inside: → GLM 5.2 - the model that shipped with better coding scores than closed frontier → KIMI K2.7 Code - long-context beast, sits at the top of open leaderboards → DeepSeek V4 Pro - reasoning at a fraction of GPT-5.6 cost → OpenAI-compatible endpoint - plug into any agent, IDE, or tool you already run Setup, 5 steps: → aiand(.)com → Sign up with Google or GitHub → If the page won't load, switch to an EU VPN and retry → Top right corner - click Credits, select $50 → In Stripe checkout, apply promo code: UNITABETAI → $50 lands. Start firing prompts. Balance drains slowly. I've been running agents on it for days and barely dented it. The promo code is valid as of the time of publication Bookmark this before they cap the promo.

Atenov int.

14,846 views • 3 months ago