Mistral: Saba
Mistral model for multilingual chat, reasoning, and tool-assisted workflows
Context
32,768
Max output
26,214
Input / 1M
$0.20
Output / 1M
$0.59
Providers
3
01Profile
Profile
Released
2025-02-17
Last updated
2025-02-17
Knowledge cutoff
—
Weights
Closed
02Capabilities
Capabilities
Input
text · pdf
Output
text
Reasoning
No
Tool calling
Yes
Structured output
Yes
Attachments
Yes
Cache read
—
Cache write
—
03Pricing
Pricing across 3 providers
Per 1M tokens, USD. Sorted by input price. First-party row highlighted.
04History
Change history
7+ recent events since 2025-02-17
2026-09-03Contextnano-gptContext window32,000→ 32,768
2026-09-03Contextnano-gptMax output32,768→ 26,214
2026-08-25ContextkiloMax output32,768→ 26,214
2026-08-25ContextopenrouterMax output32,768→ 26,214
2026-05-15Addedopenrouteroffering→ new
2026-03-15Addedkilooffering→ new
2026-02-17Addednano-gptoffering→ new