Mistral API Providers
Who serves Mistral, at what price and speed.
Mistral AI publishes open weights for its smaller models (Mistral Small, Ministral, Devstral, Mistral Nemo), which DeepInfra and others serve next to Mistral's own API.
DeepInfra served 78% of the 323B Mistral tokens that went through a large public LLM router over Oct 3 to Oct 9, 2026. Mistral AI's own API served 8%; 7 providers list the family.
Who serves Mistral tokens
All Mistral models combined, Oct 3 to Oct 9, 2026.
Mistral Nemo: prices and speed by provider
Output prices run from $0.024 to $0.15 per million tokens across 5 providers, median $0.03.
| Provider | Share | Input $/M | Output $/M | Cache read $/M | Tokens/s | Latency | Uptime 3d | Context |
|---|---|---|---|---|---|---|---|---|
| DeepInfra | 91.7% | $0.029 | $0.03 | – | 21 | 0.94s | 100.0% | 131K |
| DekaLLM | 5.2% | $0.018 | $0.03 | – | 19 | 0.73s | 100.0% | 131K |
| io.net | 3.1% | $0.018 | $0.024 | $0.012 | 43 | 0.31s | 100.0% | 128K |
| Parasail | - | $0.03 | $0.03 | – | 103 | 0.42s | 100.0% | 131K |
| Mistral | - | $0.15 | $0.15 | $0.015 | 48 | 0.51s | 100.0% | 131K |
List prices as of October 9, 2026, cheapest endpoint per provider. Tokens/s and latency are medians over 30 minutes.
Mistral Small 3.2 24B: prices and speed by provider
Output prices run from $0.20 to $0.30 per million tokens across 4 providers, median $0.275.
| Provider | Share | Input $/M | Output $/M | Cache read $/M | Tokens/s | Latency | Uptime 3d | Context |
|---|---|---|---|---|---|---|---|---|
| Venice | 100.0% | $0.094 | $0.25 | – | 30 | 1.09s | 99.8% | 256K |
| DeepInfra | - | $0.075 | $0.20 | – | 31 | 0.90s | 100.0% | 128K |
| Parasail | - | $0.09 | $0.30 | $0.05 | 50 | 0.31s | 100.0% | 131K |
| Mistral | - | $0.10 | $0.30 | $0.01 | 147 | 0.38s | 99.4% | 33K |
List prices as of October 9, 2026, cheapest endpoint per provider. Tokens/s and latency are medians over 30 minutes.
All Mistral models
| Model | Tokens, 7d | Providers | Output $/M | Context | Released |
|---|---|---|---|---|---|
| Mistral Nemo | 246B | 5 | $0.024 to $0.15 | 131K | 2024-07-19 |
| Mistral Small 3.2 24B | 36B | 4 | $0.20 to $0.30 | 256K | 2025-06-20 |
| Mistral Small 3 | 13B | 1 | $0.08 | 33K | 2025-01-30 |
| Mistral Small 4 | 11B | 1 | $0.60 to $0.66 | 262K | 2026-03-16 |
| Ministral 3 8B 2512 | 6B | 1 | $0.15 to $0.165 | 262K | 2025-12-02 |
| Ministral 3 3B 2512 | 4B | 1 | $0.10 to $0.11 | 131K | 2025-12-02 |
| Ministral 3 14B 2512 | 4B | 1 | $0.20 to $0.22 | 262K | 2025-12-02 |
| Devstral 2 2512 | 3B | 1 | $2.00 | 262K | 2025-12-09 |
| Mistral Small 3.1 24B | 516M | 1 | $0.555 | 128K | 2025-03-17 |
| Voxtral Small 24B 2507 | 144M | 1 | $0.30 to $0.33 | 33K | 2025-10-30 |
| Mixtral 8x22B Instruct | 46M | 1 | $6.00 to $6.60 | 66K | 2024-04-17 |
Finance your GPUs
Serving Mistral on your own GPUs? Tell us the cluster, the models you serve and the contracts behind them, and we will share it with partner funders that fit.
Prefer email? hello@amcompute.com