API pricing per million tokens
- Input
- $0.2
- per 1M, list
- Output
- $0.6
- per 1M tokens
- Cached input
- $0.02
- no batch tier listed
- Context
- 33K
- 26K max output
Where to buy it (3)
Mistral AI (list)the lab's own API$0.2 / $0.6cached $0.02
Mistral AIzdr$0.2 / $0.6cached $0.02
Mistral AIeu+10%$0.22 / $0.66cached $0.022
Through a gateway
Your own numbers100M input and 20M output tokens a month: $32.00 direct.
Helicone$32.00+0%
Kong AI Gateway$32.00+0%
LiteLLM$32.00+0%
Vercel AI Gateway$32.00+0%
AWS Bedrock$32.00+0%
Azure AI Foundry$32.00+0%
Google Vertex AI$32.00+0%
Databricks Mosaic AI$32.00+0%
Hugging Face Inference$32.00+0%
Cloudflare AI Gateway$33.60+5%
Requesty$33.60+5%
Eden AI$33.76+5.5%
OpenRouter$33.76+5.5%
Snowflake Cortex AI$38.40+20%
Martian Gatewayn/a
Portkeyn/a
TrueFoundry AI Gatewayn/a
IBM watsonx.ain/a
OCI Generative AIn/a
BigQuery MLn/a
Cloudflare Workers AIn/a
NVIDIA NIMn/a
Related models
Mistral Medium 3.5Mistral AI · 262K$1.5 / $7.5in / out per 1M$1.5$7.5n/a262KApr 30, 2026
Mistral Small 4Mistral AI · 262K$0.15 / $0.6in / out per 1M$0.15$0.6$0.015262KMar 16, 2026
Ministral 3 14B 2512Mistral AI · 262K$0.2 / $0.2in / out per 1M$0.2$0.2$0.02262KDec 2, 2025
Ministral 3 3B 2512Mistral AI · 131K$0.1 / $0.1in / out per 1M$0.1$0.1$0.01131KDec 2, 2025
DeepSeek V4.1 FlashDeepSeek · 1M$0.15 / $0.6in / out per 1M$0.15$0.6$0.0031MSep 10, 2026
Gemini 3.1 Flash LiteGoogle Gemini API · 1M$0.25 / $1.5in / out per 1M$0.25$1.5$0.0251MMay 7, 2026
Gemini 3.1 Flash Lite PreviewGoogle Gemini API · 1M$0.25 / $1.5in / out per 1M$0.25$1.5$0.0251MMar 3, 2026
Command A+Cohere · 192K$0.3 / $1.5in / out per 1M$0.3$1.5$0.15192KSep 22, 2026
About Saba
Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance.
Takes text, files in and returns text. First listed Feb 23, 2025. Blended price $0.3 per 1M tokens at 3 input to 1 output.
Questions
How much does Saba cost?
$0.2 per million input tokens and $0.6 per million output tokens, with cached input at $0.02.
What is the context window of Saba?
33K tokens, with up to 26K tokens of output.
Where is Saba cheapest?
Of 3 routes tracked, Mistral AI is cheapest at $0.2 input and $0.6 output per million tokens.
List prices from providers' pages on the date shown; your contract may differ. Logos via logo.dev; trademarks belong to their owners.