How to use Mistral API Key for AI chat
Ann Ng
Mistral AIAbout Mistral AI
Latest models available:
- Mistral Medium 3.1: Frontier-class multimodal model
- Magistral Medium 1.2: Frontier-class reasoning model with vision
- Mistral Small 3.2: Efficient open-source model
- Codestral 2508: Specialized for coding tasks
Key features:
- 128K+ token context window
- Function calling capabilities
- Strong multilingual support
- Competitive pricing
- Vision capabilities
- Reasoning models
- Open-source options
- European AI development
Step by step guide to use Mistral API Key to chat with AI
1. Get Your Mistral API Key
First, you'll need to obtain an API key from Mistral. This key allows you to access their AI models directly and pay only for what you use.
- Visit Mistral's API console
- Sign up or log in to your account
- Navigate to the API keys section
- Generate a new API key (copy it immediately as some providers only show it once)
- Save your API key in a secure password manager or encrypted note
2. Connect Your Mistral API Key on TypingMind
Once you have your Mistral API key, connecting it to TypingMind to chat with AI is straightforward:
- Open TypingMind in your browser
- Click the "Settings" icon (gear symbol)
- Navigate to "Models" section
- Click "Add Custom Model"
- Fill in the model information:Name:
mistral-medium-latest via Mistral(or your preferred name)Endpoint:https://api.mistral.ai/v1/chat/completionsModel ID:mistral-medium-latestfor example (check Mistral model list)Context Length: Enter the model's context window (e.g., 32000 for mistral-medium-latest)
mistral-medium-latesthttps://api.mistral.ai/v1/chat/completionsmistral-medium-latest via Mistralhttps://www.typingmind.com/model-logo.webp32000 - Add custom headers by clicking "Add Custom Headers" in the Advanced Settings section:Authorization:
Bearer <MISTRAL_API_KEY>:X-Title:typingmind.comHTTP-Referer:https://www.typingmind.com - Enable "Support Plugins (via OpenAI Functions)" if the model supports the "functions" or "tool_calls" parameter, or enable "Support OpenAI Vision" if the model supports vision.
- Click "Test" to verify the configuration
- If you see "Nice, the endpoint is working!", click "Add Model"
3. Start Chatting with Mistral models
Now you can start chatting with Mistral AI models through TypingMind:
- Select your preferred Mistral model from the model dropdown menu
- Start typing your message in the chat input
- Enjoy faster responses and better features than the official interface
- Switch between different AI models as needed
Mistral Medium 3.1- Use specific, detailed prompts for better responses (How to use Prompt Library)
- Create AI agents with custom instructions for repeated tasks (How to create AI Agents)
- Use plugins to extend Mistral capabilities (How to use plugins)
- Upload documents and images directly to chat for AI analysis and discussion (Chat with documents)
4. Monitor Your AI Usage and Costs
One of the biggest advantages of using API keys with TypingMind is cost transparency and control. Unlike fixed subscriptions, you pay only for what you actually use. Visit https://admin.mistral.ai/organization/usage to monitor your Mistral API usage and set spending limits.
| Feature | Mistral Subscription Plans | Using Mistral API Keys |
|---|---|---|
| Cost Structure | ❌ Fixed monthly fee Pay even if you don't use it Le Chat Pro:$14.99/month Le Chat Team:$24.99/user/month | ✅ Pay only for actual usage $0 when you don't use it |
| Usage Limits | ❌ Hard daily/hourly caps You have to wait for the next period to use it again | ✅ Unlimited usage No limits. Only limited by your budget |
| Model Access | ❌ Platform decides available models Old models get discontinued | ✅ Access to all API models Including older & specialized versions |
- Use less expensive models for simple tasks
- Keep prompts concise but specific to reduce token usage
- Use TypingMind's prompt caching to reduce repeat costs (How to enable prompt caching)
- Using RAG (retrieval-augmented generation) for large documents to reduce repeat costs (How to use RAG)
Access Shisa AI: Shisa V2 Llama 3.3 70B via OpenRouter
Shisa V2 Llama 3.3 70B is a bilingual Japanese-English chat model fine-tuned by Shisa.AI on Meta’s Llama-3.3-70B-Instruct base. It prioritizes Japanese language performance while retaining strong English capabilities. The model was optimized entirely through post-training, using a refined mix of supervised fine-tuning (SFT) and DPO datasets including regenerated ShareGPT-style data, translation tasks, roleplaying conversations, and instruction-following prompts. Unlike earlier Shisa releases, this version avoids tokenizer modifications or extended pretraining. Shisa V2 70B achieves leading Japanese task performance across a wide range of custom and public benchmarks, including JA MT Bench, ELYZA 100, and Rakuda. It supports a 128K token context length and integrates smoothly with inference frameworks like vLLM and SGLang. While it inherits safety characteristics from its base model, no additional alignment was applied. The model is intended for high-performance bilingual chat, instruction following, and translation tasks across JA/EN.
Access OpenAI: GPT-4.1 via OpenRouter
GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and GPT-4.5 across coding (54.6% SWE-bench Verified), instruction compliance (87.4% IFEval), and multimodal understanding benchmarks. It is tuned for precise code diffs, agent reliability, and high recall in large document contexts, making it ideal for agents, IDE tooling, and enterprise knowledge retrieval.
Access OpenAI: GPT-4.1 Mini via OpenRouter
GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.
Access OpenAI: GPT-4.1 Nano via OpenRouter
For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.
Access EleutherAI: Llemma 7b via OpenRouter
Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.
Access AlfredPros: CodeLLaMa 7B Instruct Solidity via OpenRouter
A finetuned 7 billion parameters Code LLaMA - Instruct model to generate Solidity smart contract using 4-bit QLoRA finetuning provided by PEFT library.
Access ArliAI: QwQ 32B RpR v1 (free) via OpenRouter
QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself. The model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.
Access ArliAI: QwQ 32B RpR v1 via OpenRouter
QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself. The model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.
Access Agentica: Deepcoder 14B Preview (free) via OpenRouter
DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini
Access Agentica: Deepcoder 14B Preview via OpenRouter
DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini
Access MoonshotAI: Kimi VL A3B Thinking (free) via OpenRouter
Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.
Access MoonshotAI: Kimi VL A3B Thinking via OpenRouter
Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.
