SambaNova: Models Intelligence, Performance & Price

SambaNova
SambaNova

This analysis is intended to support you in choosing the best model provided by SambaNova for your use-case.

Most Intelligent

#1
GLM-5.3 (max)GLM-5.3 (max)
32
#2
MiniMax-M3.1-FlashMiniMax-M3.1-Flash
29
#3
GLM-5.3 (low)GLM-5.3 (low)
23
#4
Gemma 4 31BGemma 4 31B
4
#5
gpt-oss-120b (high)gpt-oss-120b (high)
4

Intelligence Index

Total 8 models

Lowest Time per Task

#1
gpt-oss-120b (high)gpt-oss-120b (high)
1.9 min
#2
GLM-5.3 (low)GLM-5.3 (low)
7.2 min
#3
GLM-5.3 (max)GLM-5.3 (max)
14.2 min
#4
MiniMax-M3.1-FlashMiniMax-M3.1-Flash
15.4 min

Time per Intelligence Index task (minutes)

Total 8 models

Lowest Price

#1
MiniMax-M3.1-FlashMiniMax-M3.1-Flash
$0.20
#2
gpt-oss-120b (high)gpt-oss-120b (high)
$0.26
#3
gpt-oss-120b (low)gpt-oss-120b (low)
$0.26
#4
Gemma 4 31BGemma 4 31B
$0.46
#5
Gemma 4 31B (non-reasoning)Gemma 4 31B (non-reasoning)
$0.46

Blended price (per 1M tokens)

Total 8 models

SambaNova offers 8 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.

  • For intelligence, the top models on SambaNova are GLM-5.3 (max) (32), MiniMax-M3.1-Flash (29), and GLM-5.3 (low) (23).
  • For time per task, gpt-oss-120b (high) (1.9 min), GLM-5.3 (low) (7.2 min), and GLM-5.3 (max) (14.2 min) finish an Intelligence Index task fastest on SambaNova. Time per task varies significantly across models, with the slowest taking 8.1x longer than the quickest.
  • For latency, Llama 3.3 70B (1.25s), gpt-oss-120b (low) (1.52s), and GLM-5.3 (low) (2.01s) offer the lowest time to first answer token.
  • For pricing, MiniMax-M3.1-Flash ($0.20), gpt-oss-120b (high) ($0.26), and gpt-oss-120b (low) ($0.26) offer the lowest blended prices per 1M tokens. Prices vary up to 2.3x across models.
  • For context window size, GLM-5.3 (max) (1M), MiniMax-M3.1-Flash (1M), and GLM-5.3 (low) (1M) support the largest context windows on SambaNova.
Artificial Analysis Intelligence Index · Higher is better
Inference time (minutes) per Intelligence Index task · Lower is better
USD per 1M tokens (blended) · Lower is better

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode-AA 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1

Evaluations

Evaluations measured independently by Artificial Analysis · AA-PersonalityBench scores are percentile ranks from 1 to 100
See more

Agentic knowledge work, Elo

Agentic SaaS workflows

Agentic scientific research workflows in a terminal

Agentic software engineering in open-source repositories

Reasoning & knowledge

Professional document reasoning, All-pass

Agentic coding & terminal use

Physics reasoning

Quantitative analysis on spreadsheets & documents

Legal agentic work, Hallucination-Gated All-Pass Rate

Agentic business operations

Agentic real-world work tasks, Elo

Agentic interactive 3D experiences, Elo

Product and brand videos designed in code, Elo

Kubernetes incident root-cause analysis

Medical long context reasoning

Visual reasoning

Rich, varied, well-crafted language

Well-structured, scannable, answer-first

Truthful, acknowledges uncertainty, non-sycophantic

Use of filler and generic language

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line

Pricing

Cost per Intelligence Index Task

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line

Speed

Average of the single-request and multi-turn agentic workloads measured

Time per Intelligence Index Task

Output tokens divided by end-to-end output speed (minutes per Intelligence Index task) · Lower is better

Quality

Endpoint Accuracy Index

Each endpoint's accuracy as a % of its model's reference endpoint · Color shows whether the gap is statistically significant

Latency

Average of the single-request and multi-turn agentic workloads measured

Time to First Answer Token

Seconds until the first answer token after any reasoning · Average of the single-request and multi-turn agentic workloads measured · Lower is better

Reliability

Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured

Endpoint Error Rate

Share of requests the endpoint failed with an error, a rate limit or a timeout · Across all the single-request and multi-turn agentic workloads measured · Lower is better

Caching

Cache Hit Rate

Median across multi-turn agentic session runs in the last four weeks, updated Oct 26, 2026 · Higher is better

Cost per Task vs. Cache Hit Rate

Weighted average cost (USD) per Intelligence Index task · Cache hit rate: median of the last four weeks, updated Oct 26, 2026
Most attractive quadrant
Pareto line

Further Analysis
Z AI logo
GLM-5.3 (max)
32
$14.80
15.3
38.36
7.7%
1M
Open
MiniMax logo
MiniMax-M3.1-Flash
29
$2.97
17.2
32.13
5.6%
1M
Open
Z AI logo
GLM-5.3 (low)
23
$4.54
8.5
4.84
0.0%
1M
Open
Google logo
Gemma 4 31B
4
--
--
10.93
1.7%
131k
Open
OpenAI logo
gpt-oss-120b (high)
4
$0.49
3.0
6.22
0.0%
131k
Open
DeepSeek logo
DeepSeek V3.1 Terminus
--
--
--
--
--
131k
Open
DeepSeek logo
DeepSeek R1 Distill Llama 70B
--
--
--
--
--
131k
Open
Meta logo
Llama 3.3 70B
--
--
--
1.21
3.4%
128k
Open
Alibaba logo
Qwen3 32B
--
--
--
--
--
33k
Open
Alibaba logo
Qwen3 32B (non-reasoning)
--
--
--
--
--
33k
Open
OpenAI logo
gpt-oss-120b (low)
--
--
--
2.15
1.9%
131k
Open
Google logo
Gemma 4 31B (non-reasoning)
--
--
--
8.21
2.7%
131k
Open
DeepSeek logo
DeepSeek V3.1 Terminus (non-reasoning)
--
--
--
--
--
128k
Open

Key definitions

Frequently Asked Questions

Common questions about SambaNova