Before we dive in, let's review the current status of models:

As models from China leads open-source intelligence we expect Llama-4, and and open-source model from OpenAI to be contenders by the end of Q2'25.

Now, let's follow up on main themes from our blog in 2024:

• The last couple of months have confirmed: intelligence is getting too cheap to ignore.

2025 is shaping up to be a tipping point—smaller, highly capable models are evolving fast, and with new GPUs, edge accelerators, and open inference frameworks, the cost of running AI continues to plummet.

Open-Source AI: The Acceleration is Real

Nowhere is this shift more obvious than in the open-source space. High-quality models like Deepseek, Qwen, and LLaMA 3 are not only catching up to closed models—they’re beating them on key tasks while being more efficient and widely available. New training techniques, quantization methods (like GGUF and AWQ), and optimized inference engines (like vLLM, llama.cpp, and TGI) are making it trivial to deploy these models locally or on modest cloud instances.

Community efforts like OpenDevin, OpenAgents, and AutoGen Studio are also pushing forward the agentic paradigm—where models don’t just generate text, but act, reason, and collaborate across tools and environments.

• More Agents and Smarter Ones

Agentic flows have continued to mature rapidly in early 2025. What started with simple task chaining has evolved into full-blown multi-agent systems capable of reasoning, planning, and collaborating across complex workflows. From customer support and web search to autonomous coding assistants and data pipelines, agents are becoming more modular, more capable, and easier to orchestrate.

One major enabler of this growth has been the emergence of flexible, open frameworks like AutoGen, OpenDevin, CrewAI, and OpenAgents. These ecosystems are not just about chaining prompts—they enable persistent memory, tool use, collaborative reasoning, and API integration at scale.

But perhaps the most exciting development is the rise of the Model Context Protocol (MCP). MCP provides a standardized way to connect models to external tools, APIs, memory stores, file systems, and even live data streams—bringing structure and reliability to agentic flows. Think of it as the glue that lets multiple LLMs and tools operate in sync, with context-awareness and shared state.

• Multi-agent collaboration is now practical thanks to shared context and coordination layers.
• Tool use and web interaction have become smoother via standardized protocol interfaces.
• Long-term memory and world modeling are more achievable with persistent context layers.

With MCP and agentic frameworks working together, we’re entering a phase where AI systems don’t just respond—they operate autonomously and adaptively in open-ended environments.

We will be releasing various MCP servers so make sure check out our Github repo

• Do I believe what I see?

The lines between AI and reality is keep getting harder to distinguish, and there is not doubt that these technologies will redefine industries like entertainment, design, and marketing for better or worse!

2025 is, as expected, on track to be the year of locally deployed open-source models and the rise of agentic flows. The question is—are you ready?

Show less

One standout achievement was OpenAI's release of its first reasoning model, o1, which demonstrated remarkable proficiency in domains like mathematics, coding, and science, showcasing the next evolution of intelligent systems.

Meanwhile, smaller models have risen to prominence, matching or even exceeding the performance of previous-generation models that were 10x larger. This leap in efficiency is a game-changer, as it significantly reduces the cost per million tokens, making AI more accessible and scalable for real-world applications.

However, we are particularly excited about the ever-growing and advancing open-source ecosystem. Open-source models are increasingly capable of competing with closed-state-of-the-art (SOTA) models, while frameworks are enabling innovations like fine-tuning, quantization for efficient inference, retrieval-augmented generation (RAG), and agent flows. These advancements are democratizing access to powerful AI tools and making cutting-edge capabilities more widely available.

We’ll dive deeper into these topics later, but it's worth noting that the rumors of pre-training scaling hitting a wall could have serious implications for the field. Until now, the prevailing industry assumption has been that more parameters, more data, and more compute would inevitably lead to increasingly better models. If this paradigm is shifting, the open-source ecosystem will likely emerge as one of the winners, thanks to its focus on innovation and a “do more with less” attitude as we see in below table:

Looking ahead here is what we expect for 2025:

• Intelligence will be too cheap!

2025 will be an exciting shift in the AI landscape. Smaller, highly capable models are evolving rapidly, and with the arrival of new GPUs and inference chips, the costs are dropping faster than ever. We see that 1M tokens is now drifting down to just $0.1 for mainstream models, and renting a powerful H100 GPU can now cost less than $1/hour. Let's not forget, we’ve got Apple’s M4 chip already making waves, the RTX50 series on the horizon, and Intel Gaudi 3 cards promising to bring affordable private AI servers within everyone’s reach.

• I need more agents!

This year, we’ve seen agentic flows starting to mature, especially in areas like customer support, coding, and web search. With the availability of more multimodal models, advanced reasoning capabilities, and long-context window models, alongside powerful open-source agent frameworks, we’re excited to see how agent flows will evolve. We expect them to grow significantly and eventually replace more traditional RAG applications.

• Do I believe what I see?

While waiting for OpenAI’s SORA release, multiple companies have introduced realistic AI image and video generation models, with some even releasing open-source versions. The lines between AI and reality continue to blur, and we expect this space to witness significant advancements in both quality and accessibility. It’s becoming clear that these technologies will redefine industries like entertainment, design, and marketing in ways we’re only beginning to imagine.

2025 is shaping up to be the year of locally deployed open-source models and the rise of agentic flows. The question is—are you ready?

Show less