Side by side pricing and context for Ministral 8B and GPT-5 Nano. Rates are curated catalog values with last verified dates on each model page.
Ministral 8B
Small Mistral model with flat low rates for edge-style and high-QPS text tasks. Keep prompts short and measure monthly volume.
Input $0.150 / Output $0.150 per 1M
Context 128K · Approx
GPT-5 Nano
Cheapest GPT-5 family tier for classification, routing, and ultra-high QPS text jobs. Use when Mini is still too expensive and Exact OpenAI counts matter for budgeting.
Input $0.050 / Output $0.400 per 1M
Context 128K · Exact (o200k_base)
Quick take
- GPT-5 Nano is cheaper on input in this catalog snapshot.
- Ministral 8B is cheaper on output in this catalog snapshot.
- Both last verified on their model pages (2026-08-02 / 2026-08-02).
Rate table
| Metric | Ministral 8B | GPT-5 Nano |
|---|---|---|
| Input / 1M | $0.150 | $0.050 |
| Output / 1M | $0.150 | $0.400 |
| Cached input / 1M | n/a | $0.0050 |
| Context | 128K | 128K |
| Tokenizer | Approx | Exact (o200k_base) |