AI Tools
SambaNova DataScale
SambaNova DataScale: Preise, Fähigkeiten, Updates und Alternativen, von den Seiten des Anbieters gelesen.

Auf Englisch
Key facts
- Price
- Paid plans: see pricing Read on cloud.sambanova.ai
- Free plan
- No Read on cloud.sambanova.ai
- Platforms
- Linux · API
- API
- Not stated on the maker’s pages
- Open source
- Not stated on the maker’s pages
- Last update
- 11 Nov 2026 Read on sambanova.ai
- Prices checked
- 14 Nov 2026 Read on cloud.sambanova.ai
SambaNova DataScale is an integrated hardware and software system for organizations running AI workloads. SambaNova describes it for training and inference, including generative AI, agentic AI, and high-resolution computer vision. Deployment options described by the maker include on-premises systems and hosted data centers.
It is infrastructure, not a standalone AI chat app. SambaNova’s current product page for the DataScale URL redirects to its SambaRack page. The pricing page reviewed lists SambaCloud model-token rates, not DataScale hardware or subscription pricing, so those rates are not included here.
- Category
- Developer Tools
- Read from the maker’s pages
- 2026-11-18
Company
Features
- Runs AI model training and inference workloads
- Supports multiple models, including large open-source models
- Uses SambaNova Reconfigurable Dataflow Units and dataflow architecture
- Can be deployed on premises or in hosted data centers
- Supports high-resolution image analysis without down-sampling or tiling
- Can run chained and custom models
Use cases
- Train foundation models for enterprise or research workloads
- Deploy AI inference in an on-premises data center
- Analyze high-resolution medical, satellite, or astronomy images
- Serve multiple AI models in an organization
- Build custom computer-vision workflows for captured images
Pricing
Paid plans: see pricing
- Prices checked
- 2026-11-14
GLM-5.3
0.14 USD per 1M cached input tokens
- 0.14 USD per 1M cached input tokens
- 2.80 USD per 1M input tokens
- 8.80 USD per 1M output tokens
MiniMax-M3.1-Flash
0.03 USD per 1M cached input tokens
- 0.03 USD per 1M cached input tokens
- 0.30 USD per 1M input tokens
- 1.20 USD per 1M output tokens
gemma-4-31B-it
0.38 USD per 1M input tokens
- 0.38 USD per 1M input tokens
- 1.15 USD per 1M output tokens
gpt-oss-120b
0.22 USD per 1M input tokens
- 0.22 USD per 1M input tokens
- 0.59 USD per 1M output tokens
Meta-Llama-3.3-70B-Instruct
0.60 USD per 1M input tokens
- 0.60 USD per 1M input tokens
- 1.20 USD per 1M output tokens
Capabilities
- Choice of models — “SambaStack switches between multiple frontier-scale models, enabling complex agentic AI workflows to execute end-to-end on one node.” source
- Self-hosted — “SambaRack™ is a state-of-the-art hardware platform that can be set up easily in data centers to run Al inference workloads.” source
- API — “SambaNova provides simple-to-integrate APIs for Al inference, making it easy to onboard applications.” source
- Runs models for you — “SambaNova provides simple-to-integrate APIs for Al inference, making it easy to onboard applications.” source
- Builds agents and workflows — “Build or use pre-built AI agents — all with business-aware intelligence.” source
Latest updates
- GLM-5.3 Now Running on SambaCloud
GLM-5.3 is now available on SambaCloud at over 200 tokens/s.
- Supermicro Will Build and Sell SambaNova-Powered Inference Systems · SN50
Supermicro will deliver and support inference systems pairing NVIDIA GPUs with SambaNova SN50 RDUs, rolling out in phases.
- MiniMax M3.1 Flash Now Available on SambaCloud
MiniMax M3.1 Flash is now available on SambaCloud.
- MiniMax M3 Prompt Caching on SambaCloud: How It Works
Prompt caching is now available for MiniMax M3 on SambaCloud for shared prefixes of at least 4,096 tokens.
- MiniMax M3 Running Fastest on SambaCloud
MiniMax M3 is available on SambaCloud for developers building long-horizon, long-context agents.
Where it works
Linux · API
Alternatives to SambaNova DataScale
- Core42
Core42 provides sovereign cloud and AI infrastructure for training, deploying and running AI models, including inference through its Compass API.
- Liquid Foundation Models
Liquid Foundation Models are open-weight AI models for text, vision, audio, and decision tasks, usable locally or through Liquid's API.
- Baseten
Baseten provides infrastructure to deploy and run AI models, train models, and build agent workflows using managed inference, APIs, training tools, and sandboxes.
- NVIDIA NIM
NVIDIA NIM provides containerized, optimized inference microservices for deploying AI models on NVIDIA-accelerated infrastructure.
- Ollama
Ollama runs open AI models locally or in Ollama’s cloud, with a CLI and API for using them in apps and coding tools.
- Together AI
Together AI provides APIs and infrastructure for running, fine-tuning, and training open AI models.
More from SambaNova Systems
- SambaNova Cloud
SambaCloud is a hosted API platform for running open-source AI models, with OpenAI-compatible endpoints, model integrations, and token-based pricing.
Questions about SambaNova DataScale
Is SambaNova DataScale free?
No. The maker’s pricing page lists no free plan for SambaNova DataScale. Prices checked on 14 Nov 2026.
How much does SambaNova DataScale cost?
The plans on the maker’s pricing page: GLM-5.3 (0.14 USD per 1M cached input tokens), MiniMax-M3.1-Flash (0.03 USD per 1M cached input tokens), gemma-4-31B-it (0.38 USD per 1M input tokens), gpt-oss-120b (0.22 USD per 1M input tokens) and Meta-Llama-3.3-70B-Instruct (0.60 USD per 1M input tokens). Prices checked on 14 Nov 2026.
Which platforms does SambaNova DataScale run on?
SambaNova DataScale runs on Linux and API, as its maker’s pages state. Read on the maker’s pages on 18 Nov 2026.
Is SambaNova DataScale still available?
Yes, as of 18 Nov 2026: its maker’s pages were read that day and carried no notice that SambaNova DataScale has ended.
What changed recently in SambaNova DataScale?
The latest entry in its maker’s changelog: 11 Nov 2026, “GLM-5.3 Now Running on SambaCloud”. Before it: 3 Nov 2026, “Supermicro Will Build and Sell SambaNova-Powered Inference Systems · SN50”. Changelog checked on 14 Nov 2026.
Sources
- https://cloud.sambanova.ai/plans/pricing — Official pricing page read on 2026-11-18; it lists SambaCloud model-token rates, not DataScale pricing.
- https://sambanova.ai/hubfs/23945802/SambaNova_SambaFlow_Datasheet_English.pdf — Maker datasheet describing DataScale SN40L, its training and inference use, deployment options, and dataflow architecture.
- https://sambanova.ai/products/datascale/ — The DataScale product address redirects to the SambaRack page.
- https://sambanova.ai/products/ai-computer-vision/ — Maker page describing DataScale for high-resolution computer-vision analysis and example applications.
- https://sambanova.ai/blog/the-air-cooled-token-factory-intel-gpu-d-series-sambarack-sn50 — Maker blog post dated 2026-11-15 describing an Intel GPU D-Series and SambaRack SN50 inference demonstration.
Not confirmed
- API
- the pages reviewed do not confirm a DataScale-specific API.
- Open source
- the pages reviewed do not state whether DataScale is open source.
- Platforms
- no supported end-user platforms are specified for this hardware system.
- Latest version
- the DataScale datasheet names SN40L, but the current DataScale page redirects to SambaRack and does not confirm a current DataScale version.