Skip to main content
Source: SambaNova model docs. All model IDs use the prefix sambanova/. Powered by SambaNova RDU (Reconfigurable Dataflow Unit) chips designed for large-scale AI inference.
SambaNova models are not yet available through the Timbal platform proxy. A SAMBANOVA_API_KEY is required to use these models. If you’d like to access SambaNova models via your Timbal API key, please contact sales.

​
DeepSeek

DeepSeek-R1-0528

Reasoning · Speedsambanova/DeepSeek-R1-0528DeepSeek’s latest R1 reasoning model running on SambaNova RDU hardware. Frontier-class chain-of-thought reasoning with extended thinking support.
  • $5.00 / $7.00
  • 128K context
  • Text input
  • Thinking
  • Knowledge cutoff ~early 2025

DeepSeek-V3.2

Reasoning · Speedsambanova/DeepSeek-V3.2DeepSeek V3.2 preview running on SambaNova RDU hardware. Latest iteration of the V3 series with further capability improvements.
  • $3.00 / $4.50
  • 8K context
  • Text input
  • Knowledge cutoff ~early 2025

DeepSeek-V3.1

Reasoning · Speedsambanova/DeepSeek-V3.1DeepSeek V3.1 running on SambaNova RDU hardware. Updated general-purpose model with improved instruction-following and coding capabilities.
  • $3.00 / $4.50
  • 128K context
  • Text input
  • Knowledge cutoff ~early 2025

DeepSeek-V3.1-Terminus

Reasoning · Speedsambanova/DeepSeek-V3.1-TerminusDeepSeek V3.1 Terminus preview running on SambaNova RDU hardware. Preview release with enhanced capabilities for complex tasks.
  • $3.00 / $4.50
  • 128K context
  • Text input
  • Knowledge cutoff ~early 2025

DeepSeek-V3.1-cb

Reasoning · Speedsambanova/DeepSeek-V3.1-cbDeepSeek V3.1 community tier running on SambaNova RDU hardware at significantly reduced pricing. Ideal for high-volume workloads.
  • $0.15 / $0.75
  • 128K context
  • Text input
  • Knowledge cutoff ~early 2025

DeepSeek-V3-0324

Reasoning · Speedsambanova/DeepSeek-V3-0324DeepSeek V3 March 2024 release running on SambaNova RDU hardware. Strong general-purpose model for coding, reasoning, and instruction-following.
  • $3.00 / $4.50
  • 128K context
  • Text input
  • Knowledge cutoff ~early 2024

​
Qwen

Qwen3-235B-A22B-Instruct-2507

Reasoning · Speedsambanova/Qwen3-235B-A22B-Instruct-2507Qwen3 235B MoE model with 22B active parameters running on SambaNova RDU hardware. Supports hybrid thinking mode with strong multilingual and agentic capabilities.
  • $0.40 / $0.80
  • 64K context
  • Text input
  • Hybrid thinking
  • Knowledge cutoff ~early 2025

Qwen3-32B

Reasoning · Speedsambanova/Qwen3-32BQwen3 32B dense model running on SambaNova RDU hardware. Supports hybrid thinking mode with strong multilingual capabilities.
  • $0.40 / $0.80
  • 32K context
  • Text input
  • Hybrid thinking
  • Knowledge cutoff ~early 2025

​
Meta Llama

Llama-4-Maverick-17B-128E-Instruct

Reasoning · Speedsambanova/Llama-4-Maverick-17B-128E-InstructMeta Llama 4 Maverick preview with 17B active parameters across 128 experts running on SambaNova RDU hardware. Supports image input.
  • $0.63 / $1.80
  • 128K context
  • Text, Image input
  • Knowledge cutoff ~early 2025

Meta-Llama-3.3-70B-Instruct

Reasoning · Speedsambanova/Meta-Llama-3.3-70B-InstructMeta Llama 3.3 70B instruction-tuned model running on SambaNova RDU hardware. Multilingual, strong at coding and reasoning tasks.
  • $0.60 / $1.20
  • 128K context
  • Text input
  • Knowledge cutoff Dec 2023

Llama-3.3-Swallow-70B-Instruct-v0.4

Reasoning · Speedsambanova/Llama-3.3-Swallow-70B-Instruct-v0.4Japanese-tuned Llama 3.3 70B by TokyoTech-LLM running on SambaNova RDU hardware. Optimized for Japanese language tasks while retaining strong English capabilities.
  • $0.60 / $1.20
  • 16K context
  • Text input
  • Knowledge cutoff Dec 2023

Meta-Llama-3.1-8B-Instruct

Reasoning · Speedsambanova/Meta-Llama-3.1-8B-InstructMeta Llama 3.1 8B instruction-tuned model running on SambaNova RDU hardware. The most affordable and fastest entry point on SambaNova.
  • $0.10 / $0.20
  • 16K context
  • Text input
  • Knowledge cutoff Jul 2024

​
OpenAI OSS

gpt-oss-120b

Reasoning · Speedsambanova/gpt-oss-120bOpenAI’s open-weight MoE model with 120B total parameters running on SambaNova RDU hardware. Near-parity with o4-mini on reasoning benchmarks. Apache 2.0.
  • $0.22 / $0.59
  • 128K context
  • Text input
  • Thinking
  • Knowledge cutoff Jun 2024

​
MiniMax

MiniMax-M2.5

Reasoning · Speedsambanova/MiniMax-M2.5MiniMax M2.5 large-scale model running on SambaNova RDU hardware with an industry-leading 160K context window.
  • $0.30 / $1.20
  • 160K context
  • Text input
  • Knowledge cutoff ~early 2025