Tenstorrent
Tenstorrent builds AI training and inference accelerators on an open RISC-V architecture. It is led by Jim Keller and licenses its designs as well.
Highlights
- Wormhole and Blackhole training/inference chips on open RISC-V architecture
- Led by Jim Keller — chip designer behind Apple A4/A5, AMD Zen, Tesla FSD, Intel chip strategy
- $700M+ raised at approximately $2B valuation
- Manufactured by Samsung Foundry under multi-year contract
- Investors include Jeff Bezos, Samsung Catalyst, Hyundai, Fidelity, Eclipse Ventures
- Open-source software stack (Metalium, TT-NN, TT-Buda) under Apache license
- Targets enterprises and governments seeking NVIDIA alternatives for sovereign AI compute
External link — opens tenstorrent.com in a new tab. Tenstorrent is a third-party product; we are not affiliated with it.
Link checked 17 November 2026: this site responded and still names the product.
About Tenstorrent
What it is
Tenstorrent builds AI training and inference accelerators — the Wormhole and Blackhole chips — on an open RISC-V architecture, fabricated by Samsung Foundry. It is led by Jim Keller, whose previous work includes Apple's A4 and A5, AMD's Zen and Tesla's self-driving chip. It has raised over $700M at roughly a $2B valuation, and licenses its designs as well as selling hardware.
Why it's different
The distinguishing choice is openness. Nvidia's position rests as much on CUDA as on silicon, and every competitor has to answer the software question; Tenstorrent's answer is an open architecture and an open software stack, which lets customers license designs and build their own rather than buy a closed box. That is genuinely different positioning. It is also the harder road: the software ecosystem around CUDA represents nearly two decades of accumulated tooling, and porting real workloads to anything else remains meaningfully more work than the benchmarks suggest.
How people use it
This is bought by organisations building AI infrastructure at a scale where Nvidia's pricing and allocation are strategic problems, and by companies licensing the IP to put AI silicon in their own products. For most developers it is background rather than a purchase — but credible competition is the main force that changes what inference costs, so it is worth knowing it exists.
Written by the n3os team. We are not affiliated with Tenstorrent.
More Model Platforms & APIs tools
- Actualyze AIPaid
Actualyze AI is a single OpenAI-compatible layer between an organisation's applications and the model providers that governs and secures every request.
- AurikoFreemium
Auriko is an LLM gateway: one OpenAI-compatible endpoint in front of more than a dozen model providers, routing requests by price, latency and reliability.
- EtchedPaid
Etched builds Sohu, an application-specific integrated circuit designed to do one thing: transformer inference. It targets inference, not training.
- HirundoPaid
Hirundo removes the effect of particular training data, such as bad labels or copyrighted material, from an already-trained model without full retraining.
- Hydra HostPaid
Hydra Host rents dedicated, non-virtualised NVIDIA GPU servers for AI, HPC and data workloads, across a network of more than forty data centres.
- Inception LabsPaid
Inception Labs is a Stanford spinout building Mercury, a family of diffusion-based language models that refine the whole output in parallel.
Every Model Platforms & APIs tool in the directory →
Listed by how many tags they share with Tenstorrent, then by name. Nobody pays to appear here.
This listing was written from public information, without Tenstorrent’s involvement. If you own it and something here is wrong — or you would rather not be listed at all — email us and, once we have confirmed it is yours, we will correct or remove it.
Get the ones worth knowing about
We write one of these for every tool worth the trouble. Get the new ones, plus what we have found genuinely useful lately.
Your address goes to Buttondown, who send the emails on our behalf. One click unsubscribes, and the list is never sold or shared.