diffbot - Network

To learn more, check out the repo: https://github.com/diffbot/diffbot-llm-inference

To learn more, check out the repo: https://github.com/diffbot/diffbot-llm-inference

89,886 developers are building their own Perplexity on-prem with Diffbot LLM —https://huggingface.co/diffbot/Llama-3.1-Diffbot-Small-2412

89,886 developers are building their own Perplexity on-prem with Diffbot LLM —

https://huggingface.co/diffbot/Llama-3.1-Diffbot-Small-2412

The model isn't the moat. Perplexity can be recreated as a side project. #DeepSeek proved this. We proved this.Download Diffbot LLM. Run...

The model isn't the moat. Perplexity can be recreated as a side project. #DeepSeek proved this. We proved this.

Download Diffbot LLM. Run it off your own GPU. Congrats, your on-prem #AI is smarter than Perplexity.

1. Diffbot LLM is a side project. Sonar is Perplexity's entire business.2. We used the profits from our primary business to train...

1. Diffbot LLM is a side project. Sonar is Perplexity's entire business.

2. We used the profits from our primary business to train Diffbot LLM. Perplexity raised $915M to train theirs.

3. We open sourced Diffbot LLM. Perplexity chose to keep theirs secret.

Let's be frank here. The score difference is insignificant. And we'll probably play SimpleQA tag for awhile.What IS significant is...

Let's be frank here. The score difference is insignificant. And we'll probably play SimpleQA tag for awhile.

What IS significant is how we got here vs. Perplexity.

While working on my talk last week, Perplexity released Sonar Pro API with a special emphasis on its factuality benchmark F1 score of 0.858,...

While working on my talk last week, Perplexity released Sonar Pro API with a special emphasis on its factuality benchmark F1 score of 0.858, handily beating other internet connected LLMs like Gemini-2.0-flash.

The SimpleQA benchmark they used is open source and LLM judged, so I set it up to run the 4000 question eval on Diffbot LLM overnight and went to bed.

The next morning, we beat Sonar Pro.

Perplexity Sonar Pro API launched last week as the best performing model on factuality.24 hours later, it's the 2nd best performing...

Perplexity Sonar Pro API launched last week as the best performing model on factuality.

24 hours later, it's the 2nd best performing model (and it's not because of DeepSeek).

Our hack night last year with @neo4j and @weaviate was a blast so we’re kicking off 2025 with another!Same place, same time. We might even...

Our hack night last year with @neo4j
and @weaviate was a blast so we’re kicking off 2025 with another!

Same place, same time. We might even demo Diffbot LLM!

RSVP here: https://lu.ma/hacknight-github-01-22-25

#sanfrancisco #ai

A demo is also available at https://diffy.chat.We look forward to building a future of grounded AI with you all.

A demo is also available at https://diffy.chat.

We look forward to building a future of grounded AI with you all.

Diffbot LLM's lighter footprint puts on-prem hosting well within reach. And we are excited to share that we are releasing Diffbot LLM...

Diffbot LLM's lighter footprint puts on-prem hosting well within reach.

And we are excited to share that we are releasing Diffbot LLM open source on @github with weights available for download on #Hugginface

https://github.com/diffbot/diffbot-llm-inference

At Diffbot, we believe that general purpose reasoning will eventually be distilled down to ~1B parameters. Knowledge is best retrieved at...

At Diffbot, we believe that general purpose reasoning will eventually be distilled down to ~1B parameters.

Knowledge is best retrieved at inference, outside of model weights.

The benefit of full source attribution goes two ways. Not only is credit provided to publishers, every fact is also independently...

The benefit of full source attribution goes two ways.

Not only is credit provided to publishers, every fact is also independently verifiable.

Every response from Diffbot LLM draws from the results of real-time expert web searching and queries to the Diffbot Knowledge Graph....

Every response from Diffbot LLM draws from the results of real-time expert web searching and queries to the Diffbot Knowledge Graph.

Naturally, this means Diffbot LLM always provides full attribution to its cited sources.

We launched the world's most grounded #LLM — Diffbot #GraphRAG LLM.Instead of training on ever larger corpuses of data, Diffbot LLM...

We launched the world's most grounded #LLM — Diffbot #GraphRAG LLM.

Instead of training on ever larger corpuses of data, Diffbot LLM is trained to be an expert web researcher.

In fact, Diffbot LLM makes zero assumptions about its knowledge of the world.

Start your hunt: https://app.diffbot.com/get-startedSome notes: - Searching on the dashboard does not consume credits - Sign up for a fresh...

Start your hunt: https://app.diffbot.com/get-started

Some notes:
- Searching on the dashboard does not consume credits
- Sign up for a fresh trial if your trial has expired

🪺 We've hidden 16 easter eggs in the Diffbot Knowledge Graph. 🐣 Each egg contains a clue to the next.👀 Eggs are only visible in...

🪺 We've hidden 16 easter eggs in the Diffbot Knowledge Graph.
🐣 Each egg contains a clue to the next.
👀 Eggs are only visible in entities on the Dashboard (link in comment), not via API.
🔍 This is not school. Google/ChatGPT/query all you want.
🥇 First person to eggstract all 16 easter eggs will win a lifetime personal plan and the swaggiest Diffbot swag.

Your first clue — "The Easter Bunny was first mentioned in an essay by this German physician."

Tapping into data in the Diffbot Knowledge Graph can feel a lot like you're doing the research yourself, except in light speed.Every...

Tapping into data in the Diffbot Knowledge Graph can feel a lot like you're doing the research yourself, except in light speed.

Every fact includes a link to a public source. So you're never left wondering if you've broken a privacy law.

Frankly, I just find it fun to see where facts come from. In this LeadGraph screenshot, the KG was able to link Microsoft Teams to adidas by analyzing an employee's job description with NLP.

#nlp #knowledgegraph

Every few years or so, someone tries to reinvent Wikipedia. The honor system works for a bit, but they all inevitably run into trolls and...

Every few years or so, someone tries to reinvent Wikipedia.

The honor system works for a bit, but they all inevitably run into trolls and spammers.

The solution to more reliable facts isn't to have a single consolidated knowledge base, but a #knowledgegraph that links a single fact to several sources.

Guess I’ll have to start practicing dictating DQL… #WWDC

Guess I’ll have to start practicing dictating DQL… #WWDC

When friends ask me why fact finding AI is necessary. #JustAIThings

When friends ask me why fact finding AI is necessary. #JustAIThings