Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

Software can now understand human language and act on it, in real time. Introducing Sage2 - Extract: natural language in, structured data out. 166x cheaper than Opus 5.5, beats Cloudflare's Clairic - Decide: the first decision model with streaming. 3x faster, because it reads the context as it grows.... show more

Levanto Labs

1,593 subscribers

174,167 Aufrufe • vor 5 Stunden •via X (Twitter)

Wissenschaft & Technologie Finanzen

Anya Rossi• Live Now

Private livecam show

50 Kommentare

Profilbild von Levanto Labs

Levanto Labsvor 5 Stunden

Sign-ups in the next 24 hours get 20M free tokens: Want to give it a try? This is our playground: Learn how to set up MPP with Sage2 for your agents here:

Profilbild von Levanto Labs

Levanto Labsvor 5 Stunden

Extraction models are new. We couldn’t find a benchmark, so we made one. On our Extract Bench, Sage2 was the most accurate. Sage2 ranked first both in accuracy (77.8) and identifying non-answers (72.2) beating @cloudflare’s Clairic scored (75.3, 59). It’s open-source, we invite improvements:

Profilbild von fred

fredvor 5 Stunden

@Cloudflare beating clairic on extraction while being this cheap is a strong launch

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Let us know your thoughts on the benchmark.

Profilbild von Nick.B

Nick.Bvor 5 Stunden

@Cloudflare congrats on the launch! Sage 2 is killer

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare The extraction feature lets you pull out words from posts 👀

Profilbild von gelo

gelovor 5 Stunden

@Cloudflare That’s impressive

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare 🔥

Profilbild von Nurstar 🌟

Nurstar 🌟vor 5 Stunden

@Cloudflare Congrats team! Looking forward to playing with it!

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Aeon is going to go nuts with this.

Profilbild von Joachim | AI news

Joachim | AI newsvor 5 Stunden

@Cloudflare Small models built for one narrow job are where the real savings are. Most apps don't need a frontier model to turn an email into a clean row of data, and 166x cheaper makes that obvious.

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Smart, fast, and cheap at runtime.

Profilbild von Avid

Avidvor 4 Stunden

@Cloudflare yooo this is fire let me get that api and try it out

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare DMing you API Keys.

Profilbild von Avid

Avidvor 4 Stunden

@Cloudflare my grok bot shall get it

Profilbild von Kleanthi

Kleanthivor 5 Stunden

@Cloudflare looking forward to put the 20M tokens to work 🫡

Profilbild von Levanto Labs

Levanto Labsvor 5 Stunden

@Cloudflare 🔥

Profilbild von matt

mattvor 5 Stunden

@Cloudflare this is the kind of primitive devs actually need, not another chatbot

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Fluid intelligence.

Profilbild von aeon ⭐

aeon ⭐vor 5 Stunden

@Cloudflare Huge!

Profilbild von Levanto Labs

Levanto Labsvor 5 Stunden

@Cloudflare Thanks to your eval work, we were able to fix multiple bugs before launching!

Profilbild von marcus

marcusvor 5 Stunden

@Cloudflare Game changer??? yes

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Correct.

Profilbild von natt

nattvor 5 Stunden

@Cloudflare been waiting for someone to make decision models feel real-time. this looks like it

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Combining the extraction -> decision calls moves it into a new level of fluid intelligence at runtime.

Profilbild von Shann³

Shann³vor 5 Stunden

@Cloudflare Congrats on the launch! Excited to start building with Sage2

Profilbild von Luca

Lucavor 5 Stunden

@Cloudflare huge

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Let's get to work.

Profilbild von Zhen

Zhenvor 5 Stunden

@Cloudflare Excited that this is out!

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare What will you build?

Profilbild von zen.

zen.vor 5 Stunden

@Cloudflare let's gooo team

Profilbild von elvis

elvisvor 5 Stunden

@Cloudflare I have been testing these extraction models, and they're quite useful in many agentic workflows. It's good to see efforts to make them more affordable, so it's easier to scale.

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare Let us know what you think after trying it!

Profilbild von Shahaf Antwarg

Shahaf Antwargvor 5 Stunden

@Cloudflare looks amazing

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare 🔥

Profilbild von JustCarl (τ, τ)

JustCarl (τ, τ)vor 4 Stunden

@Cloudflare Great launch!

Profilbild von ben.

ben.vor 4 Stunden

@Cloudflare 166x cheaper than opus is a wild opening line

Profilbild von gus

gusvor 4 Stunden

@Cloudflare this goes hard as hell

Profilbild von Levanto Labs

Levanto Labsvor 4 Stunden

@Cloudflare The 3d renders were made with @fframes_rust.

Profilbild von Taka.eth

Taka.ethvor 4 Stunden

@Cloudflare huge

Profilbild von ares. 🎧

ares. 🎧vor 4 Stunden

@Cloudflare streaming decisions is the part nobody's talking about yet

Profilbild von Francesco Andreoli ᵍᵐ

Francesco Andreoli ᵍᵐvor 4 Stunden

@Cloudflare the soundtrack music is just epic Chris

Profilbild von Chimp of the North

Chimp of the Northvor 4 Stunden

@Cloudflare Congrats on the launch 👏🏻

Profilbild von gg

ggvor 4 Stunden

@Cloudflare congrats on the launch!!!

Profilbild von Daniele Chillè

Daniele Chillèvor 4 Stunden

@Cloudflare Go go go!

Profilbild von Serge Dynnyk

Serge Dynnykvor 4 Stunden

@Cloudflare Congrats on the launch, team, looking forward to seeing more from you!

Profilbild von roman

romanvor 4 Stunden

@Cloudflare Sage 2 is cool af! Congrats for your work!

Profilbild von Daniel Priscu

Daniel Priscuvor 4 Stunden

@Cloudflare streaming decisions matter most for voice. waiting for the full sentence kills the latency

Profilbild von Chen

Chenvor 4 Stunden

@Cloudflare 166x cheaper than Opus 5.5 is such a wild gap i almost had to double check the number

Profilbild von Tempo

Tempovor 5 Stunden

@Cloudflare making it easy for agents to use Sage 2

Ähnliche Videos

Introducing streaming payments. Track tokens in real-time with Metronome. Collect payments in real-time with Tempo. Get paid for every token as it’s burned.

Introducing streaming payments. Track tokens in real-time with Metronome. Collect payments in real-time with Tempo. Get paid for every token as it’s burned.

Stripe

116,516 Aufrufe • vor 6 Monaten

Introducing Higgsfield x Claude Opus 5.5. Anthropic’s newest model, which is 40% cheaper and over 30% faster than Opus 5, is live on Higgsfield. Pair Opus 5.5 with the Higgsfield MCP and turn your most ambitious ideas into complete products. Available on Claude via Higgsfield MCP and in Higgsfield Supercomputer.

Introducing Higgsfield x Claude Opus 5.5. Anthropic’s newest model, which is 40% cheaper and over 30% faster than Opus 5, is live on Higgsfield. Pair Opus 5.5 with the Higgsfield MCP and turn your most ambitious ideas into complete products. Available on Claude via Higgsfield MCP and in Higgsfield Supercomputer.

Higgsfield AI 🧩

2,955,974 Aufrufe • vor 1 Monat

Agent payments will soon overtake human payments on the internet. The Machine Payments Protocol (Machine Payments Protocol) is a new open standard co-authored by Stripe and Tempo. It’s designed to be extensible and payment-method agnostic, already supporting stablecoins, cards, and more.

Agent payments will soon overtake human payments on the internet. The Machine Payments Protocol (Machine Payments Protocol) is a new open standard co-authored by Stripe and Tempo. It’s designed to be extensible and payment-method agnostic, already supporting stablecoins, cards, and more.

Tempo

423,138 Aufrufe • vor 7 Monaten

Introducing Cognee v1.0: a major breakthrough in agentic intelligence. It is 145% better than Opus 4.8 and GPT 5.5 at long context memory retrieval. Cognee allows a 100 BILLION token context window 100,000x more than Claude. It's: - 6.9x cheaper than GPT 5.5 and Opus 4.8 - Cold starts in 350ms & searches in 260ms Why this matters: Today agents forget important context, redo tasks, waste tokens, and slow down as workflows get more complex. Cognee solves this. It’s not a place to build agents. It connects to the agents you’ve already built, across any platform, and makes them significantly cheaper, faster, and more accurate. Here's how it works:

Introducing Cognee v1.0: a major breakthrough in agentic intelligence. It is 145% better than Opus 4.8 and GPT 5.5 at long context memory retrieval. Cognee allows a 100 BILLION token context window 100,000x more than Claude. It's: - 6.9x cheaper than GPT 5.5 and Opus 4.8 - Cold starts in 350ms & searches in 260ms Why this matters: Today agents forget important context, redo tasks, waste tokens, and slow down as workflows get more complex. Cognee solves this. It’s not a place to build agents. It connects to the agents you’ve already built, across any platform, and makes them significantly cheaper, faster, and more accurate. Here's how it works:

Vasilije

849,548 Aufrufe • vor 4 Monaten

AI. RWAs. Agent-to-agent commerce. It's all converging on x402, Coinbase's protocol for machine-native payments. Now agents can pay each other in stablecoins at the HTTP layer, and they can do it in real time on Sei. Jay breaks it down with FINTECH.TV ↓

AI. RWAs. Agent-to-agent commerce. It's all converging on x402, Coinbase's protocol for machine-native payments. Now agents can pay each other in stablecoins at the HTTP layer, and they can do it in real time on Sei. Jay breaks it down with FINTECH.TV ↓

Sei

16,868 Aufrufe • vor 8 Monaten

This morning we shared Google launched AP2 (Agent Payment Protocol) with EigenCloud as a launch partner. Now see it in action: we’ve extended Google's Agent Development Kit demo to show natural-language payments + storage on EigenDA, agents can store/retrieve data, pay with crypto rails (X402), and get back proofs instantly. Agents can now store + pay on EigenDA… just by asking in plain English 🤯 Try it yourself 👇

This morning we shared Google launched AP2 (Agent Payment Protocol) with EigenCloud as a launch partner. Now see it in action: we’ve extended Google's Agent Development Kit demo to show natural-language payments + storage on EigenDA, agents can store/retrieve data, pay with crypto rails (X402), and get back proofs instantly. Agents can now store + pay on EigenDA… just by asking in plain English 🤯 Try it yourself 👇

EigenCloud

324,027 Aufrufe • vor 1 Jahr

Your agents are only as good as their context, and the most valuable context is often locked behind a login. Introducing Authenticated Fetch, 1 request to instantly extract data from logged-in sites. It’s 3x faster than a full browser session, and available on all plans today.

Your agents are only as good as their context, and the most valuable context is often locked behind a login. Introducing Authenticated Fetch, 1 request to instantly extract data from logged-in sites. It’s 3x faster than a full browser session, and available on all plans today.

Browserbase

10,542 Aufrufe • vor 6 Tagen

GraphEngine V1 is LIVE and it launches first on Base Starting today, anyone can query live onchain activity on Base using natural-language AI. No SQL. No subgraph coding. Just ask and get structured, real-time intelligence. This is the beginning of a new era for onchain insight.

GraphEngine V1 is LIVE and it launches first on Base Starting today, anyone can query live onchain activity on Base using natural-language AI. No SQL. No subgraph coding. Just ask and get structured, real-time intelligence. This is the beginning of a new era for onchain insight.

GraphAI

18,860 Aufrufe • vor 11 Monaten

Introducing Natura-1, nature will's first general foundation model for the physical world. It is a single model that follows plain-language instructions zero-shot and carries them out with dexterous hands. - 50B MoE, pretrained on 1M hours of egocentric human experience. - Drives 20-DoF dexterous hands from plain language, zero-shot, in a 100ms loop. - Scaling human data from ~44K to 1M hours cut action MSE on held-out robot data by 67.7%. - Built on in-house data and compute infra, it handles fine-grained long-horizon tasks and a ~30-minute zero-shot interaction in one continuous take. The general model exists, and it can only get stronger.

Introducing Natura-1, nature will's first general foundation model for the physical world. It is a single model that follows plain-language instructions zero-shot and carries them out with dexterous hands. - 50B MoE, pretrained on 1M hours of egocentric human experience. - Drives 20-DoF dexterous hands from plain language, zero-shot, in a 100ms loop. - Scaling human data from ~44K to 1M hours cut action MSE on held-out robot data by 67.7%. - Built on in-house data and compute infra, it handles fine-grained long-horizon tasks and a ~30-minute zero-shot interaction in one continuous take. The general model exists, and it can only get stronger.

Ning Ding

21,925 Aufrufe • vor 20 Tagen

Introducing Palmier x Claude Opus 5.5. Anthropic’s newest model, which is 40% cheaper and over 30% faster than Opus 5, is live on Palmier Pro. Pair Opus 5.5 with the Palmier MCP and turn your most ambitious video ideas into reality. Available on Claude via Palmier MCP. This is actually insane.

Introducing Palmier x Claude Opus 5.5. Anthropic’s newest model, which is 40% cheaper and over 30% faster than Opus 5, is live on Palmier Pro. Pair Opus 5.5 with the Palmier MCP and turn your most ambitious video ideas into reality. Available on Claude via Palmier MCP. This is actually insane.

Palmier

12,977 Aufrufe • vor 1 Monat

Introducing SubQ Snap, the first long-context decision model for real enterprise work. Snap is the only decision model or LLM where compute per token stays constant as context grows. • 384K tokens of multimodal context, 12x larger than Jev. • 61.7 on Decision Index v0.2.1 vs Jev’s 57.9, outperforming on 13 of 16 enterprise benchmarks. •Processes 100K tokens and returns a decision in under one second. •$0.05/MM input tokens. Output is free. Available today to anyone. The first 10,000 sign-ups get 100M free input tokens. We’ll likely hit capacity fast and will add users as we add compute. Compute offers welcome 🙂. Latency will be variable as we tune the platform to this new model and category of users. Model speeds are 80ms for minimal context, 250ms for 32k, 660ms for 100k, and 2.5 seconds for 384k. Visit:

Introducing SubQ Snap, the first long-context decision model for real enterprise work. Snap is the only decision model or LLM where compute per token stays constant as context grows. • 384K tokens of multimodal context, 12x larger than Jev. • 61.7 on Decision Index v0.2.1 vs Jev’s 57.9, outperforming on 13 of 16 enterprise benchmarks. •Processes 100K tokens and returns a decision in under one second. •$0.05/MM input tokens. Output is free. Available today to anyone. The first 10,000 sign-ups get 100M free input tokens. We’ll likely hit capacity fast and will add users as we add compute. Compute offers welcome 🙂. Latency will be variable as we tune the platform to this new model and category of users. Model speeds are 80ms for minimal context, 250ms for 32k, 660ms for 100k, and 2.5 seconds for 384k. Visit:

Alexander Whedon

4,513,405 Aufrufe • vor 21 Tagen

Overshoot is the fastest API for real-time vision. Literally faster than human reaction time (<200ms). Developers are using it to ship video agents in gaming, robotics, sports and security. Check out the playground: Congrats Zakaria and Younes on the launch!

Overshoot is the fastest API for real-time vision. Literally faster than human reaction time (<200ms). Developers are using it to ship video agents in gaming, robotics, sports and security. Check out the playground: Congrats Zakaria and Younes on the launch!

Y Combinator

51,517 Aufrufe • vor 8 Monaten

OpenAI just announced "GPT-4o". It can reason with voice, vision, and text. The model is 2x faster, 50% cheaper, and has 5x higher rate limit than GPT-4 Turbo. It will be available for free users and via the API. The voice model can even pick up on emotion and generate emotive voice.

OpenAI just announced "GPT-4o". It can reason with voice, vision, and text. The model is 2x faster, 50% cheaper, and has 5x higher rate limit than GPT-4 Turbo. It will be available for free users and via the API. The voice model can even pick up on emotion and generate emotive voice.

Lior Alexander

485,137 Aufrufe • vor 2 Jahren

So far, machine payments have mostly been about businesses accepting stablecoins. And stablecoins are great (cross-border, low cost, etc.)! But much of the planet is holding a card. 🤖 💳 🌎 Here’s a 📺 sneak peek demo from Steve Kaliski of how a business can programmatically accept cards via agents. As an example: Stripe Climate is a way to contribute to carbon removal funding. Stripe Climate implemented the Machine Payments Protocol using the Stripe API, so agents can make micropayment donations ( In addition to accepting stablecoin (specifically usdc on Tempo), agents can now pay with fiat methods, namely cards and Link (and soon buy-now-pay-laters) via Shared Payment Tokens (SPTs), using the same Machine Payments Protocol integration. Human buyers, their agents, and businesses each have their own preferred methods of payment. Humans often already have a card. Agents may prefer stablecoins. Businesses tend to want to accept any form of payment, as long as it is high conversion, low fraud, and properly priced. For microtransactions, stablecoins are a low-cost way (no fixed card fee) to transact. But if you want to tell your agent to donate $100 on your behalf, a card may be the preferred solution. If you’re a business and want to accept machine payments, whether over stablecoins, cards, Link, or other future methods, read about Shared Payment Tokens ( and sign up for our machine payments private preview: And stay tuned: more to make it easy for consumers, agents, and businesses to transact, shipping over the next few weeks. (And excuse the acronyms and specifics terms, there's a lot of new infrastructure and primitives being built to support agentic commerce.)

So far, machine payments have mostly been about businesses accepting stablecoins. And stablecoins are great (cross-border, low cost, etc.)! But much of the planet is holding a card. 🤖 💳 🌎 Here’s a 📺 sneak peek demo from Steve Kaliski of how a business can programmatically accept cards via agents. As an example: Stripe Climate is a way to contribute to carbon removal funding. Stripe Climate implemented the Machine Payments Protocol using the Stripe API, so agents can make micropayment donations ( In addition to accepting stablecoin (specifically usdc on Tempo), agents can now pay with fiat methods, namely cards and Link (and soon buy-now-pay-laters) via Shared Payment Tokens (SPTs), using the same Machine Payments Protocol integration. Human buyers, their agents, and businesses each have their own preferred methods of payment. Humans often already have a card. Agents may prefer stablecoins. Businesses tend to want to accept any form of payment, as long as it is high conversion, low fraud, and properly priced. For microtransactions, stablecoins are a low-cost way (no fixed card fee) to transact. But if you want to tell your agent to donate $100 on your behalf, a card may be the preferred solution. If you’re a business and want to accept machine payments, whether over stablecoins, cards, Link, or other future methods, read about Shared Payment Tokens ( and sign up for our machine payments private preview: And stay tuned: more to make it easy for consumers, agents, and businesses to transact, shipping over the next few weeks. (And excuse the acronyms and specifics terms, there's a lot of new infrastructure and primitives being built to support agentic commerce.)

Jeff Weinstein

21,635 Aufrufe • vor 6 Monaten

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

Claude

29,472,686 Aufrufe • vor 1 Monat

Real-time transcription just got a significant upgrade. Universal-3-Pro is now available for streaming — bringing AssemblyAI's most accurate speech model to live audio for the first time. Developers building voice agents, live captioning tools, and real-time analytics pipelines now get three things they've been asking for: 🔹 Best-in-class word error and entity detection across streaming ASR benchmarks 🔹 Real-time speaker labels — know who said what, as it happens 🔹 Superior entity detection for names, places, orgs, and specialized terminology in real-time 🔹 Code-switching and global language coverage built-in

Real-time transcription just got a significant upgrade. Universal-3-Pro is now available for streaming — bringing AssemblyAI's most accurate speech model to live audio for the first time. Developers building voice agents, live captioning tools, and real-time analytics pipelines now get three things they've been asking for: 🔹 Best-in-class word error and entity detection across streaming ASR benchmarks 🔹 Real-time speaker labels — know who said what, as it happens 🔹 Superior entity detection for names, places, orgs, and specialized terminology in real-time 🔹 Code-switching and global language coverage built-in

AssemblyAI

15,016 Aufrufe • vor 8 Monaten

Introducing Claude Opus 5.5! 40% lower cost than Opus 5 with cache reads 60% cheaper. It performs at the level of Claude Fable 5.1 for most tasks. I celebrated with a pelican riding a bike in Three.js.

Introducing Claude Opus 5.5! 40% lower cost than Opus 5 with cache reads 60% cheaper. It performs at the level of Claude Fable 5.1 for most tasks. I celebrated with a pelican riding a bike in Three.js.

Addy Osmani

419,644 Aufrufe • vor 1 Monat

.OpenAI GPT-5.5 is now available on Databricks, with Codex coding workflows and model inference fully governed through Unity AI Gateway. With GPT-5.5 on Databricks, you can: - Power coding workflows with Codex or other coding agents - Build custom agents grounded in enterprise data - Ask business questions in natural language from enterprise data with Genie - Automate document intelligence pipelines with Lakeflow Spark Declarative Pipelines Hear more from our co-founder, Patrick Wendell, and OpenAI CRO Denise Holland Dresser about how Databricks and OpenAI are bringing frontier AI to the enterprise. Get started today:

.OpenAI GPT-5.5 is now available on Databricks, with Codex coding workflows and model inference fully governed through Unity AI Gateway. With GPT-5.5 on Databricks, you can: - Power coding workflows with Codex or other coding agents - Build custom agents grounded in enterprise data - Ask business questions in natural language from enterprise data with Genie - Automate document intelligence pipelines with Lakeflow Spark Declarative Pipelines Hear more from our co-founder, Patrick Wendell, and OpenAI CRO Denise Holland Dresser about how Databricks and OpenAI are bringing frontier AI to the enterprise. Get started today:

Databricks

63,309 Aufrufe • vor 6 Monaten