Recommended Free Tools
Sakana AI announced a $30 million seed round on January 16, 2024, led by Lux Capital. The Tokyo startup was not simply trying to shrink a conventional large language model. Its broader plan was to use evolution, specialization and cooperation among models to build capable AI with potentially less data, training expense or computing infrastructure than a single giant system.
What Sakana announced in January 2024
Sakana AI, founded in Tokyo in 2023, said it had raised $30 million in seed funding. Lux Capital led the round. Participants included Khosla Ventures, 500 Global, Miyako Capital, Basis Set Ventures, JAFCO, July Fund, Geodesic Capital, Learn Capital, NTT Group, KDDI CVC and Sony Group. Individual backers named in announcements included Google AI researcher Jeff Dean, Scale AI founder Alexandr Wang and Hugging Face co-founder Clément Delangue.
The company said the money would fund a research lab in Japan, hiring, work with Japanese technology and cloud partners, and exploration of Asian markets. Contemporary coverage described an announcement-era team of about 10 people; that estimate should not be treated as Sakana’s current headcount.
Some secondary reports placed the company’s post-money valuation at roughly $200 million. Sakana’s own seed announcement did not state a valuation, so the figure remains an attributed estimate rather than an official company disclosure.
#1 Best Overall
- High-Performance AI Processor：The MS-02 Ultra features an Intel Core Ultra 9 285HX (24C/24T, up to 5.5 GHz, 13 TOPS NPU), delivering fast and efficient performance for AI inference, algorithm development, and media workloads. A PCIe x16 expansion slot supports desktop-class GPU upgrades for advanced model training and accelerated computing tasks. It's ideal for creators, engineers, and teams handling intensive parallel workloads.
- 4 × M.2 PCIe 4.0 + 4 × DDR5 SODIMM slots:Four DDR5 SODIMM slots support up to 256 GB of memory, while ECC helps maintain data integrity in mission-critical environments. Four PCIe 4.0 M.2 slots support up to 24 TB of storage, supporting RAID 0/1/5/10, combining high-speed performance with data protection. It allows for the creation of independent scratch disks, media libraries, and project drives, providing high-throughput for production workflows.
- PCIe & USB 4.0 v2: Up to three PCIe slots can be equipped, including a dual-slot x16 GPU. The main slot supports PCIe 5.0, meeting the needs of high-bandwidth creative and computing workloads. USB 4.0 v2 (80Gbps) supports high-bandwidth external storage and displays.
- Ultra-fast Networking: Wi-Fi 7 further enhances wireless performance with next-generation speeds and low-latency stability. Intelligent bandwidth switching optimizes throughput in different network environments, ensuring optimal performance for enterprise or local networks. Dual 25GbE ports (providing up to approximately 3.125 GB/s bandwidth, about 25 times faster than traditional 1GbE), enabling seamless large-scale file transfers and parallel computing. 10GbE and 2.5GbE ports, with support for Intel vPro technology, ensure enterprise-grade remote management and deployment flexibility.
- Server-grade thermal architecture: Utilizing a dedicated CPU/GPU airflow design, equipped with a 6-pipe dual-fan cooler, it maintains stable performance even under sustained loads, delivering up to 140W Turbo power while maintaining a 100W TDP, and operating with noise levels as low as 36 dB. An integrated 350W power supply ensures stable and reliable output for demanding computing tasks and fully loaded extended configurations.
Sakana’s announcement and Lux Capital’s investment note describe the round and the company’s mission.
“Smaller models” was shorthand for a wider technical strategy
The headline framing was directionally right but incomplete. Sakana described its goal as building “nature-inspired foundation models” based on evolution and collective intelligence. Schools of fish and flocks of birds supplied the analogy: many relatively simple elements can cooperate, specialize and adapt without being one centrally designed organism.
In engineering terms, Sakana was asking whether useful intelligence could come from combining specialized models, routing work among agents and automatically searching for effective combinations. That differs from the dominant scaling recipe of training one increasingly large model on more parameters, data and compute.
What the approach could involve
- Specialization: use models tuned for a language, domain or task rather than one universal system for every request.
- Collaboration: let several models or agents contribute to a result.
- Evolutionary search: use optimization to discover useful model, layer or weight combinations instead of relying only on manual experimentation.
- System-level efficiency: seek savings through routing, reuse and sparsity, not merely by deleting parameters.
That distinction matters. A smaller model can reduce latency or inference cost, but a collection of models may create orchestration, memory and monitoring overhead. “Smaller” also does not automatically mean safer, more reliable or capable of matching a frontier model.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Why investors saw an opportunity
AI development was becoming more capital-intensive as companies raced to train and operate larger systems. A successful alternative could potentially lower inference costs, run on more constrained infrastructure, adapt better to local languages and industries, and reduce dependence on a single general-purpose provider.
Japan was part of the investment thesis, not just the company’s address. Japanese telecom and technology investors could provide enterprise relationships, cloud and infrastructure partnerships, and access to demand for Japanese-language and culturally relevant AI. Lux framed Sakana as a Japan-based research company exploring alternatives to established Transformer scaling and initially targeting Asian markets.
Those points explain why the round was attractive; they do not prove that Sakana’s methods would be cheaper or better in production. The seed financing was a bet on a research direction.
The founders and the Japan connection
The announcement-era leadership included CEO David Ha, formerly associated with Google Brain and other AI research, CTO Llion Jones, a co-author of the 2017 Transformer paper while at Google Research, and Ren Ito, who later became chairman and had been associated with Mercari and Japan’s Ministry of Foreign Affairs. Early staff also came from Google, Google DeepMind, Preferred Networks, Stability AI, Rinna and Japanese research institutions.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Sakana’s current company information lists Ha, Ito and Jones as its founders. That wording is more precise than labeling every early contributor a co-founder.
How model collaboration differs from other efficiency techniques
Sakana’s work is often described loosely as “making smaller models,” but several distinct methods are involved:
Rank #3
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
| Method | What it does |
|---|---|
| Model merging | Combines existing models or their components to seek a model with useful combined abilities. |
| Model compression | Reduces a model’s size or computational burden. |
| Distillation | Trains a smaller model to reproduce behavior learned by a larger model. |
| Mixture-of-experts or orchestration | Routes requests or subtasks among specialized components. |
These approaches can overlap in a system, but they are not interchangeable. Merging an existing set of models is not the same as training a new model from a massive dataset, and a low parameter count does not by itself establish lower total operating cost.
What Sakana built after the seed round
Evolutionary Model Merge
In March 2024, Sakana introduced Evolutionary Model Merge. The method uses evolutionary optimization to search for ways of combining open-source models rather than asking researchers to choose every merge manually. The objective is to discover task-specific combinations that might inherit useful capabilities from their source models.
This is model-combination research, not ordinary compression. Its practical results still depend on evaluation, licensing and the behavior of the source models. Sakana later said related work was adopted in open-source tooling and accepted by Nature Machine Intelligence.
The AI Scientist
Sakana also released The AI Scientist, with open-source code and a paper describing a workflow for generating research ideas, running experiments and drafting papers. It is best understood as an autonomous or semi-autonomous research workflow, not a replacement for scientific judgment.
Generated experiments and text require checking, reproduction and peer review. “Fully automated scientific discovery” describes the project’s ambition and architecture; it is not proof that the system independently produces reliable science in every field.
Rank #4
- FAST RUNS IN THE FAMILY — The 16-inch MacBook Pro with the M5 Pro or M5 Max chip brings next-generation speed and powerful on-device AI to personal, professional, and creative tasks. With all-day battery life, double the starting storage,* and a breathtaking Liquid Retina XDR display, it’s pro in every way.*
- BUCKLE UP — Along with a next-generation CPU, faster unified memory, and up to 2x faster SSD storage,* M5 Pro and M5 Max feature a more powerful GPU with a Neural Accelerator built into each core, delivering faster AI performance and on-device training capabilities. So you can blaze through demanding workloads at mind-bending speeds.
- BUILT FOR AI — Apple silicon, and every major component that powers it, is designed to run demanding on-device AI workloads like LLM inference and training. And Apple Intelligence helps you write, express yourself, and get things done effortlessly with groundbreaking privacy protections at every step.*
- ALL-DAY BATTERY LIFE — MacBook Pro delivers the same exceptional performance whether it’s running on battery or plugged in.*
- MACOS RUNS APPS FAST — All your go-to apps run lightning fast in macOS, including built-in apps like FaceTime and Messages. Plus, built-in virus protection and free software updates help keep your Mac running smoothly and securely.
Later financing and a broader company
Sakana announced an approximately $200 million Series A in September 2024. Its later Series B announcement describes 32 billion yen, or approximately $200 million, using the company’s stated currency conversion and timing. Round labels, exchange rates and announcement dates can affect how those amounts are reported, so they should not be combined into one cumulative figure without a defined basis.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11The company’s current positioning is broader than the 2024 “smaller models” shorthand. Its company information page lists The AI Scientist, multi-agent orchestration foundation models, Namazu LLMs for Japan and the Darwin Gödel Machine, alongside user-facing products called Sakana Chat, Sakana Marlin and Sakana Fugu.
What has—and has not—been demonstrated
Sakana’s work supports the claim that it is pursuing model evolution, merging, orchestration and automated AI research. The available announcements do not establish that its systems are universally cheaper than OpenAI, Google or Anthropic models, match frontier-model quality, consume less energy in production, run locally on consumer hardware, or have broad commercial adoption.
There are also practical trade-offs:
- Routing mistakes can undermine a multi-model system.
- Loading several models can raise aggregate memory use.
- Specialized models may fail outside their target domain.
- Merging can create unexpected behavior or degrade capabilities.
- Open-source components may carry incompatible licenses or uncertain training-data provenance.
- Several sequential calls can make a “small” system expensive or slow.
- Smaller systems can still hallucinate, leak data and behave unsafely.
Enterprise buyers may also prefer one supported API over a stack that requires custom evaluation, compute, monitoring and incident response.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Why Sakana matters beyond one funding round
Sakana represents a challenge to the assumption that progress must come mainly from ever-larger centralized models. If its research direction succeeds, AI developers could combine reusable open models, specialize systems for local languages and workflows, and automate more of the model-development process.
Best Value
- 【High-Performance APU】The MS-S1 MAX features an AMD Ryzen AI Max+ 395 APU, integrating a Zen 5 architecture CPU (up to 5.1GHz, 16C/32T, 64M L3 Cache), an RDNA 3.5 GPU, and an NPU (50 TOPS). The total system output is 126 TOPS. It provides powerful parallel computing capabilities for demanding AI workflows. It is ideal for running local LLMs, multimodal models, and computationally intensive tasks
- 【128GB UMA Memory】Equipped with up to 128GB of LPDDR5x-8000MT/s unified memory, it enables the CPU and GPU to access a shared, high-bandwidth memory pool with extremely low latency. Ideal for large-scale AI inference, 3D workloads, and complex timelines in video editing. It eliminates traditional VRAM bottlenecks, ensuring smoother data transfer during high-intensity computations. The UMA design maximizes performance stability under high loads
- 【Flexible Expansion】The MS-S1 MAX features USB4 V2 (up to 80Gbps), dual 10GbE LAN, HDMI 2.1 (up to 8K60), a full-length PCIe x16 expansion slot, and dual M.2 slots supporting up to 16TB RAID 0/1. Wi-Fi 7 provides stronger signal coverage and a more stable wireless experience. The slide-out design facilitates upgrades and maintenance. It easily adapts to personal, studio, or rack-mount enterprise environments
- 【High-Efficiency Cooling System】Utilizing an aerospace-grade aluminum alloy chassis, copper base plate, six heat pipes, dual turbine fans, and advanced PCM thermal conductive material, it maintains stable cooling performance even under continuous load. This system supports 130W continuous power and 160W peak power operation, with a built-in 320W power supply. It boasts multiple global certifications including CCC, FCC, UL, CE, and UKCA, ensuring stable and reliable operation in various environments
- 【Cluster Design】Two MS-S1 MAX units can be configured as a dual-unit cluster to run a large 235B Q4 model locally, achieving an output speed of 10.87 tok/s. Supporting 2U rack deployment, multiple MS-S1 MAX units can be cascaded into a distributed cluster to create a high-efficiency AI computing center. A cluster of four MS-S1 MAX units successfully ran a DeepSeek-R1 671B Q4 large model. A reserved cluster power-on interface allows for unified start-up and shutdown
That possibility is especially relevant to Japan’s AI ecosystem, where domestic infrastructure, Japanese-language performance and partnerships with established technology companies are strategic concerns. It is also relevant to enterprise teams weighing hosted general-purpose APIs against systems they can adapt and operate themselves.
The defensible conclusion is narrower than “small models will replace large models.” Sakana’s seed round funded an experiment in making AI development more efficient through evolution, specialization, cooperation and automated discovery. The company’s subsequent research and financing show that this became a broader frontier-AI program, not a single product promise.
Frequently Asked Questions
Was Sakana’s $200 million valuation officially confirmed?
No. Secondary reports placed the seed-stage post-money valuation at roughly $200 million, but Sakana’s January 2024 funding announcement did not state a valuation.
Does Sakana sell a publicly priced small language model?
The company lists Sakana Chat, Sakana Marlin and Sakana Fugu, but the cited public materials do not establish consumer pricing, enterprise plans or a simple downloadable replacement for a frontier hosted model.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsQuick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
