← All models
Balanced — strong quality at a mid priceOpen Weights
Saba Pricing
Mistral AI · mistralai-mistral-saba
Input / 1M tokens
$0.200
Output / 1M tokens
$0.600
Context window
32,768 tokens
≈ 44 pages of text
Max output
—
What it costs in practice
Typical request (1,200 in + 400 out tokens)$0.0005
1,000 requests / month$0.48/mo10,000 requests / month$4.80/mo100,000 requests / month$48.00/moEstimate your own workloadOpens the calculator with Saba preloaded — adjust volume and token counts there.
Price history
Only one price point recorded so far.
The staircase chart appears once a price change is detected.
What it excels at
Low-cost text inference with 32k context; suitable for high-volume production workloads where open weights enable self-hosting or fine-tuning.
The business tradeoff
Quality trails current closed flagships on complex reasoning; output quality and latency vary with deployment hardware and optimization.
Cheaper from Mistral AI
Head-to-head
Vendor list rates, as of Jun 15, 2026 · source: openrouter · per-request examples assume 1,200 input + 400 output tokens.