Providers
Mistral
European AI with Mistral Large, Medium, and Small models
🟣
Mistral
European AI - Mistral Large, Medium, Small
Mistral
French AI company with excellent models. Good for European data residency requirements.
Setup
1. Get API Key
Get your API key from console.mistral.ai
2. Add Environment Variable
# .env.local
MISTRAL_API_KEY=...3. Configure Provider
<YourGPTProvider
runtimeUrl="/api/chat"
llm={{
provider: 'mistral',
model: 'mistral-large-latest',
}}
>
<CopilotChat />
</YourGPTProvider>Available Models
| Model | Context | Best For |
|---|---|---|
mistral-large-latest | 128K | Complex reasoning |
mistral-medium-latest | 32K | Balanced |
mistral-small-latest | 32K | Fast, efficient |
codestral-latest | 32K | Code generation |
open-mixtral-8x22b | 64K | Open-weight flagship |
Recommended: mistral-large-latest for quality, mistral-small-latest for cost efficiency.
Configuration Options
llm={{
provider: 'mistral',
model: 'mistral-large-latest',
temperature: 0.7,
maxTokens: 4096,
topP: 1,
}}European Data Residency
Mistral is based in France and offers European data processing:
// Good for GDPR compliance
<YourGPTProvider
llm={{
provider: 'mistral',
model: 'mistral-large-latest',
}}
// Data stays in Europe
>If you need European data residency for compliance, Mistral is an excellent choice.
Code Generation (Codestral)
Mistral's code-specialized model:
<YourGPTProvider
llm={{
provider: 'mistral',
model: 'codestral-latest',
}}
systemPrompt="You are a coding assistant. Help users write and debug code."
>Tool Calling
useToolWithSchema({
name: 'translate_text',
description: 'Translate text between languages',
schema: z.object({
text: z.string(),
from: z.string(),
to: z.string(),
}),
handler: async ({ text, from, to }) => {
// Mistral handles European languages very well
const translated = await translate(text, from, to);
return { success: true, data: { translated } };
},
});Pricing
| Model | Input | Output |
|---|---|---|
| mistral-large | $2/1M tokens | $6/1M tokens |
| mistral-small | $0.2/1M tokens | $0.6/1M tokens |
| codestral | $0.2/1M tokens | $0.6/1M tokens |
Check Mistral pricing for current rates.
Next Steps
- Azure - Enterprise OpenAI
- Custom Provider - Build your own