PRODUCTS

How Cohere's small model solves one big problem

Cohere is the latest lab to bet on the benefits of niche models. 

On Thursday, the company debuted Cohere Parse, a vision language model specifically designed to process large volumes of enterprise documents in a cost-effective manner. Parse can convert complex, multi-modal documents into machine-readable data that enterprises can leverage more easily for AI use-cases. 

Cohere's Parse closes a gap between expensive frontier vision models and less expensive (and less accurate) traditional optical character recognition models, the company said.

• Cohere Parse runs at $1.50 per 1,000 pages, compared to models from hyperscalers that cost roughly $10 per 1,000 pages.

• According to Cohere, Parse offers the best price-performance tradeoff of the models it evaluated, which included models from OpenAI, Anthropic, Databricks, Mistral, Microsoft, Google and Amazon. 

• The model is also available in nine languages, and will expand to meet customer demand. 

Kyle Lastovica, a spokesperson for Cohere, told The Deep View that Parse targets a very specific bottleneck for enterprises that's becoming critical in the age of AI adoption, especially as companies work to more deeply integrate agents and retrieval-augmented generation (RAG) into their workflows. 

"The quality of what goes into those systems matters more than ever, and enterprises are increasingly paying attention to AI spending," said Lastovica. "Standard optical character recognition misses key ingredients to document comprehensibility like tables, diagrams and visual structure. Many companies currently address this problem via frontier models, driving spend sharply upward."

Additionally, Lastovica said, the model can be deployed in a private cloud or in an on-premise environment with a small footprint, which is particularly important for customers in regulated industries like finance or healthcare.

Cohere Parse provides the latest of many examples of companies developing AI for specific tasks and seeing impressive results. Though document parsing may not be the flashiest or most exciting use case, it's a practical one that a lot of enterprises use AI for on a regular basis. It's the same thesis that drove NextLM's recent research in purpose-built, efficient, and affordable models for sales tasks. After months of shelling out on expensive token bills, some companies are becoming weary over bulky, frontier models and jack-of-all-trades AI. As a result, niche models, domain-specific models, and small models are gaining traction.

TOGETHER WITH CELONIS

Don't bet your business on a fortune cookie

Eliminate operational blind spots. Give AI the context it needs to succeed.

AI models are brilliant, but they’re probabilistic. Without a ground truth of your operational reality, trusting AI is like trusting a fortune cookie.

The Celonis Context Model provides this operational reality through a dynamic, real-time, system-agnostic digital twin of your operations. It translates your business into a language that AI understands.

Give your AI the context it needs to drive meaningful ROI. Meet the Context Model.

HARDWARE

Inside Agility's bet on factory-first humanoids

Humanoid robots continue to capture the imagination of humans, and the recent AI revolution has only accelerated the momentum. 

But despite demonstrations of robotics tidying up homes or awkwardly dancing up stages, there may be a better place to start: Factory floors, Pras Velagapudi, CTO of Agility Robotics, told The Deep View. With a significant labor shortage and an abundance of repetitive and dangerous tasks, manufacturing may be the perfect proving ground for humanoid robots. It's why Agility has focused on this niche, and already has robots deployed with companies such as Toyota, Amazon, and Mercado Libre. 

Velagapudi sat down with The Deep View in August at Ai4 in Las Vegas to discuss how humanoids can solve the "islands of automation," how these robots could shift the blue-collar labor force, and the path to robots eventually becoming household tools. This interview has been edited for brevity and clarity. 

Rubio-Licht: Something that I've been drilling down into the more I research physical AI is the humanoid form factor. Why do you think this is the right bet? 
Velagapudi: We have a long history of trying to get robots out in the world. Now, what's accelerated in the past few years is that we've kind of cracked all of the core technologies necessary to do this. We had been working for a long time on the physical hardware, the controls necessary to do that, and we had been making good progress and basically got all those pieces together. And then the physical AI boom happened, and has been carrying us on this wave of capability that, combined with all the hardware that we've been able to put together, is really making it so that we feel like this is a compelling time for this particular platform.

Rubio-Licht: And why are humanoids right for manufacturing versus, say, robotic arms?
Velagapudi: It isn't really humanoids versus robotic arms. It's more that manufacturing is a huge industry. One of the reasons we're going after it is because it's such a huge industry that even with robot arms and (Autonomous Mobile Robots), ground robots moving around and welding robots and all of these different pieces, there's still so much that hasn't been automated yet. What we call that is islands of automation. That there are these pieces of automation that are then separated by some sort of manual process, and then another piece of automation. But there are these little islands that exist that need to be connected by some manual process. And so there's still so much of that that's been difficult to automate because it has the wrong structure for a robot arm, or it has to occur in many different places in a facility. 

Rubio-Licht: How do you think robotics are going to eventually impact the labor force? 
Velagapudi: What we're taking on are really the dull, dirty, dangerous jobs — the 3D's of robotics — and we're very focused on areas like bulk material handling, picking up 50-pound totes, putting them into loading and unloading machinery. This is in hot environments, in dangerous environments where there's risk of repetitive strain, where you're working around other machinery. So what we're doing with these deployments is really trying to get those jobs filled — the ones that are this type of manual labor that's not really well-suited to humans.

Rubio-Licht: Do you anticipate Agility expanding into the household domain? 
Velagapudi: It's definitely in the future. We do want to get there. We do think that one of the great untapped markets for humanoid robots is closest to humans, like in homes, in hospitals, care facilities, things like that. It's where there's a lot of value because you have the most human-centric environment. There, you have the benefit of creating a robot that can do all the things in your environment that you do in your environment, but the road there to do it safely is going to be a bit longer. 

Click here to read the full interview on The Deep View.

TOGETHER WITH CELONIS

Context doesn't just inform AI. It puts AI to work.

Ask a generic AI model about outstanding invoices, and you'll get a well-worded, generic answer. Ask Enterprise AI grounded in your operational reality, and it knows which invoices are overdue, why and what to do next – then acts on it across your systems.

That's what the Celonis Context Model makes possible: a living digital twin of your business that gives AI the operational clarity to decide and act, not just answer.

No more isolated chatbots guessing at outcomes. Just decisions, grounded in reality, that move through your operations on their own.

Meet the Context Model.

PRODUCTS

As models commoditize, Replit embraces routing

State-of-the-art AI is a moving target. Replit wants to move with it. 

On Thursday, the AI coding platform launched Intelligent Model Routing as the default for all accounts. This system matches tasks to the model they're best suited for, aiming to balance quality, speed and cost. 

In its blog post, Replit said that Intelligent Model Routing offers power when a task demands it and efficiency when it doesn't. 

• And the company has already seen significant results: The company said that Intelligent Model Routing offered the same output quality as its previous version of Max Mode, the company's top-tier, paid setting that uses frontier models, at a 65% lower cost. 

• Additionally, users will still have transparency about when they are using certain models: Users in Free Mode, its unpaid tier, will be notified when Replit wants to escalate their work to higher-powered models that can incur usage costs, and can choose to override this. Paid users can also still manually select models.

• And for enterprises that want more control over which models their teams use, administrators can define an approved set of models, and Replit's model routing service will choose from that set for a given task. 

"Users can focus on how AI can augment their work, instead of model management," the company said in its blog post. 

Replit's offering follows a similar development from Snowflake, which debuted its own dynamic model routing solution within Cortex AI Gateway and its flagship AI products, Snowflake CoCo and Snowflake CoWork, in mid-August. 

Both of these firms target an increasingly popular niche in enterprise AI: Companies are growing more interested in picking the right model for the job, rather than the most powerful model on the market.

It's clear that the AI tides have shifted in recent months away from power-at-all-costs towards efficiency. This shift is showing up in numerous ways, including model routing services like Replit's, growing research and offerings around task-specific and niche models, and even new architectures that make small models as powerful as trillion-parameter ones. This change, however, leaves me with one major question: If the major model labs become commoditized, what happens to the economics of an industry built on promises of perpetually growing AI spend?