Mistral AI
Mistral Large 4 (Le Chonk)
Mistral Large 4 (Le Chonk) public preview: text and image input via Vercel AI Gateway, with a 524,288-token Gateway context cap.
Try it on magicdoor.aiPublic preview. Availability depends on the magicdoor.ai model picker and the Mistral endpoint on Vercel AI Gateway; no separate Mistral account is needed.
Try Mistral Large 4 (Le Chonk) without stacking subscriptions
magicdoor.ai includes all 14 chat models in one workspace. Basic is $6/month with $2 usage balance in the first month, then $1 on renewal; additional use is usage-based.
Capabilities
Vision
Web Search
Image Generation
Canvas
Reasoning
Code Interpreter
PDF Support
Pricing on magicdoor.ai
+ usage-based pricing
- • $0.68 per 1M input tokens
- • $2.09 per 1M output tokens
- • October 8, 2026 Gateway sale rates: $0.68 input, $0.07 eligible cached input, $2.09 output per 1M tokens; rates may change
- • Gateway context: 524,288 tokens; maximum completion: 262,144 tokens (Mistral lists 1M context for direct access)
- • The Magicdoor route requires per-request ZDR and pins Mistral; requests fail closed if the endpoint or guarded sale rates change
- • No GreenPT renewable-inference route, PDF input, code interpreter, or explicit reasoning-effort control is exposed for this model
- • Basic includes $2 usage balance in the first month, then $1 on renewal
- • No separate Mistral AI subscription needed
- • Access all 14 chat models in one workspace
Sources (checked October 8, 2026):
What Mistral Large 4 (Le Chonk) Is Best For
Text and image analysis
Compare responses on a grounded prompt with text and an image; this endpoint returns text.
Long-context drafts
Work with larger text inputs while respecting the 524,288-token Gateway context ceiling.
Structured tasks
Test tool use and structured output where supported by the Magicdoor integration.
Second opinion
Switch to another available model and check a real task against your own criteria.
How It Works
Start using Mistral Large 4 (Le Chonk) in under a minute
Sign up for magicdoor.ai ($6/mo)
Select Mistral Large 4 (Le Chonk) from the model picker
Start chatting — pay only for what you use
Keep Comparing Before You Subscribe
Mistral Large 4 (Le Chonk): route and sale pricing
Check official sources, the Gateway cap, guarded sale rates, and provider-inference privacy scope.
Which AI model should I use?
Answer a few questions and get a model recommendation.
How to save money on AI
Use cheaper models first, then switch up when the task needs it.
Frequently Asked Questions
Is Mistral Large 4 generally available?
No. Mistral calls Large 4 (Le Chonk) a public preview, not GA. Check the magicdoor.ai picker for current availability.
What does Mistral Large 4 cost on magicdoor.ai?
As checked October 8, 2026, Vercel AI Gateway lists sale rates of $0.68 per 1M input tokens, $0.07 per 1M eligible cached input tokens, and $2.09 per 1M output tokens. Sale rates can change; check current model pricing.
Is Large 4 a one-million-token model on this route?
Mistral lists 1M context for direct access, but the Vercel AI Gateway endpoint on this route lists 524,288 tokens of context and 262,144 maximum output tokens.
What privacy route does Magicdoor use for Large 4?
The Magicdoor route pins Mistral on Vercel AI Gateway, requires per-request ZDR, and fails closed if the guarded endpoint changes or becomes unavailable. This provider-inference policy is separate from magicdoor.ai chat storage. No GreenPT route is listed.
Try Mistral Large 4 (Le Chonk) on magicdoor.ai
All major AI models in one place — no separate subscriptions needed.
Open your doorAlready have an account? Sign in
