The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →PewDiePie’s Ajax is reported as a fine-tuned Qwen3.5-9B model intended to operate inside Odysseus, a self-hosted workspace for chat, agents, research, documents, email, notes and calendars. The project is designed around local control, but as of October 3, 2026, the available coverage does not establish a public download, hardware requirements, benchmark results or a license for Ajax’s model weights.
The creator also says OpenAI deactivated his account twice while he was developing Ajax. That account history comes from PewDiePie and an email he displayed; OpenAI has not provided a specific explanation in the reporting reviewed.
What Ajax is supposed to be
Tom’s Hardware reported on October 2, 2026, that PewDiePie introduced Ajax as a “fine-tuned Qwen3.5-9B model.” The “9B” label identifies the model’s parameter class; it is not a performance score or proof of a particular speed or quality level.
Ajax is intended for Odysseus, a self-hosted AI workspace. The project’s description covers conversational chat, agents, research, documents, email, notes, calendar functions and local-model workflows. The Ajax page presents the system as an always-on assistant for activities such as web search, browsing, email and calendar work. Those privacy and capability statements are product claims, not independent security or performance tests.
#1 Best Overall
- Supercomputer performance directly to your desk in a compact, energy-efficient design, enabling enterprise-scale AI and high-performance computing right where you need it.
- The power of Grace Blackwell architecture, delivering up to 1 petaFLOP of AI performance for local model fine-tuning, inference, and analytics, accelerating your time-to-solution.
- Designed from the ground up to build and run AI, delivering seamless integration of the full NVIDIA AI software stack —so you can develop locally and deploy anywhere.
- NVIDIA DGX Spark gives you the freedom to experiment, prototype, and innovate faster by augmenting laptop, desktop, cloud, or data center resources. With more power to learn, prototype, test, and innovate, NVIDIA DGX Spark delivers exceptional ROI for increased productivity.
- Use NVIDIA DGX Spark to unlock new ideas and experiment with large models (up to 200 billion parameters at FP4) directly on your desktop with 128GB of unified memory. Empower rapid testing, validation, and iteration—driving innovation in a secure, high-performance setting.
Why PewDiePie says OpenAI deactivated his account
According to Tom’s Hardware’s account of PewDiePie’s video, OpenAI deactivated his account twice during Ajax development. An email shown in the video reportedly cited “distillation” for one deactivation. The account was restored once, then deactivated again after he used a model to create seed data. The displayed message reportedly did not give specific examples of the conduct that triggered the action.
In this context, distillation means using outputs, answers or reasoning from one model to help train another model. PewDiePie said he wanted to “distill just a little bit” to improve Ajax. His description should not be treated as an independently confirmed account of OpenAI’s enforcement decision: the reviewed reporting does not include a detailed OpenAI statement about these two actions.
What “uncensored” means here
The launch coverage says Ajax’s refusal behavior was modified with automatic abliteration using Heretic, an open-source tool. In practical terms, that means the project attempted to reduce or remove some of the model’s built-in refusal responses; it does not demonstrate that every safeguard was removed or that the model will answer every prohibited request.
Rank #2
- Built for Local AI Development: AMD Ryzen AI Halo is designed for local AI development and inference, featuring 128GB unified memory and support for up to 200B parameter models to build and run intensive AI workloads locally.
- 128GB Unified Memory: Features 128GB LPDDR5x unified memory at 8000 MT/s with 256 GB/s memory bandwidth, providing a shared memory pool across the CPU, GPU, and NPU to support larger AI models.
- AMD Ryzen AI Max+ 395 Processor: Features 16 cores, 32 threads, and Zen 5 architecture, paired with AMD Radeon 8060S integrated graphics featuring 40 RDNA 3.5 compute units and an AMD XDNA 2 NPU with up to 50 TOPS.
- Linux AI Developer Platform: Purpose-built for Linux-based AI development with full AMD ROCm software support and preloaded tools, models, and workflows optimized for local AI development.
- Compact, Connected Design: Includes a 2TB M.2 SSD, 10GbE LAN, Wi-Fi 7, Bluetooth 5.4, USB-C connectivity, and HDMI 2.1b.
PewDiePie said he drew a boundary around harming other people or oneself. On legal advice, he described Ajax as “not designed to provide dangerous actionable instructions.” That is an intended safety position, not an independent safety evaluation. No published testing in the reviewed sources establishes how Ajax behaves on dangerous, illegal, self-harm or privacy-sensitive prompts.
Further reinforcement learning, another decensoring step, quantization and benchmarking were described as future work for a later version. They should not be presented as completed features of the announced model.
Can you download Ajax yet?
Availability was unresolved in the launch coverage. A Free Press Journal report published October 2 said Ajax was not then available for download and that a final model card, public benchmark, confirmed weights license and other release details were absent. Other reports used terms such as “unveiled” or “released,” but those words do not by themselves establish that downloadable weights exist.
Rank #3
- EVOLUTION AMD RYZEN AI MAX+ 395 MINI PC - GMKtec EVO-X2 is the next evolution in AI mini PC Ryzen Strix Halo series. Thanks to AMD Simultaneous Multithreading (SMT) the core-count is effectively doubled, to 32 threads. Ryzen AI Max+ 395 has 64 MB of L3 cache and can boost up to 5.1 GHz, depending on the workload. The Ryzen AI Max+ 395 is currently rated as the "most powerful x86 APU" on the market for AI computing.
- AI NPU with XDNA 2 ARCHITECTURE - Powered by 16 “Zen 5” CPU cores, 50+ peak AI TOPS XDNA 2 NPU and a truly massive integrated GPU driven by 40 AMD RDNA 3.5 CUs, the Ryzen AI MAX+ 395 is a transformative upgrade and delivers a significant performance boost over the competition. The Ryzen AI Max+ 395 excels in consumer AI workloads like the llama.cpp-powered application: LM Studio. Shaping up to be the must-have app for client LLM workloads, LM Studio allows users to locally run the latest language model without any technical knowledge required and unleash their creativity and productivity.
- AMD RADEON 8090S iGPU GAMING PC - The AMD Radeon RX 8060S offers all 40 CUs with up to 2.9 GHz graphics clock and uses the new RDNA 3.5 architecture. The powerful iGPU is positioned between an RTX 4060 and 4070 laptop GPU and therefore enables gaming in FHD at maximum details in most demanding games. The 8060S can also utilize the full 64GB pool, which is perfect for running LLMs such as Deepseek 32B, which runs comfortably on this machine.
- EIGHT CHANNEL LPDDR5X - LPDDR5X is a new ground breaking memory small form factor installed on-board. With blazing speeds up to to 8000MT/s, it runs 1.5x faster than the DDR5 SODIMMs; 90% better performance over DDR5 SODIMMs in video conferencing and photo editing; 30% better performance in productivity apps; 4% better performance in digital content workloads.
- QUAD SCREEN 8K DISPLAY SUPPORT - EVO-X2 AI Mini PC support 4-screen 4K/8K output via HDMI 2.1 (8K@60Hz), DisplayPort 1.4 (4K@60Hz), and dual USB 4 40Gbps Transfer speed (supporting PD3.0/DP1.4/DATA). Ideal for gaming, video editing, and multitasking, it provides expansive and crisp multi-display support.
| Question | Status reported for October 2–3, 2026 |
|---|---|
| Public Ajax download | Not conclusively established; check the official project page before attempting installation. |
| Final model card | Not reported as available. |
| Ajax weights license | Not established by the available sources. |
| Benchmark results | No public performance figures were identified in the launch coverage. |
| Hardware requirements | Not stated. The Qwen3.5-9B designation alone is not a system requirement. |
| Planned next-version work | Reinforcement learning, additional decensoring, quantization and benchmarking were described as planned. |
Odysseus licensing is not Ajax licensing
The Odysseus repository describes the workspace software as licensed under AGPL-3.0-or-later. That governs the repository’s software, not necessarily the Ajax model files. Until the project publishes a weights license, users should not assume that AGPL permissions apply to Ajax or that commercial redistribution is allowed.
Odysseus is self-hosted, so its design points toward running services under the user’s control rather than relying solely on a hosted chatbot. However, no source reviewed for this announcement specifies a processor, graphics card, memory capacity, operating system or quantized package that would confirm “home PC” compatibility.
What to verify before treating Ajax as a usable local model
When an official release appears, check these items separately:
- Weights and files: confirm that the download is an official Ajax release and identify the exact model revision.
- Model card: look for training-data notes, intended use, limitations, safety testing and known failure modes.
- License: read the model-weights terms independently from Odysseus’s AGPL software license.
- Hardware guidance: use the publisher’s memory and accelerator requirements for the exact precision or quantization, rather than inferring them from “9B.”
- Reproducible benchmarks: distinguish creator-supplied results from independent tests and note the hardware and settings used.
- Safety behavior: test refusal consistency and unsafe-output handling only within lawful, controlled evaluations; “less restricted” is not a safety score.
Bottom line
Ajax is best understood as a reported Qwen3.5-9B fine-tune aimed at the self-hosted Odysseus workspace, accompanied by PewDiePie’s account of two OpenAI deactivations linked to model distillation. The “uncensored” label refers to an attempted alteration of refusal behavior, not verified removal of all safeguards. Public availability, licensing, benchmarks and PC requirements remained unconfirmed in the reporting available on October 3, 2026.
Quick 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.
