Mistral Token Counter
Current Mistral models use the Tekken tokenizer, a tiktoken-based BPE trained on more than 100 languages. TokenMath counts Mistral tokens with a calibrated estimate, typically within ±3%.
Universal AI Token Counter
Zero-latency, client-side BPE tokenization with real-time multi-model benchmarks.
Mistral Models: Context Windows & Token Prices
| Model | Context Window | Input / 1M | Output / 1M | Token Count Accuracy |
|---|---|---|---|---|
| Mistral Large 3 | 256k tokens | $0.5 | $1.5 | Calibrated Mistral Tokenizer (±3%) |
| Mistral Small 4 | 256k tokens | $0.15 | $0.6 | Calibrated Mistral Tokenizer (±3%) |
| Codestral (Current Generation) | 256k tokens | $0.3 | $0.9 | Heuristic Estimate (±5%) |
How Mistral Tokenization Works
Mistral models split text with the Tekken tokenizer, a Tekken BPE (tiktoken-based) scheme with a ~131,000 tokens vocabulary. Common words map to one token, while rare words, identifiers and non-Latin scripts are split into several subword pieces. Read how BPE and SentencePiece tokenization work for the details.
- Tekken compresses source code and many European languages more efficiently than older SentencePiece-based Mistral tokenizers.
- Older Mistral models used a 32k SentencePiece vocabulary and produce noticeably more tokens for the same text.
Once you know your token count, estimate a full monthly bill with the LLM cost calculator, or see every Mistral AI model price.
Mistral Token Counter FAQ
How do I count Mistral tokens?
Paste your prompt into the counter above with a Mistral model selected. TokenMath tokenizes the text in your browser and shows the token count, context window usage and input cost. Accuracy: Calibrated Mistral Tokenizer (±3%).
What tokenizer does Mistral use?
Mistral models are counted with the Tekken tokenizer (Tekken BPE (tiktoken-based), ~131,000 tokens vocabulary). Current Mistral models use the Tekken tokenizer, a tiktoken-based BPE trained on more than 100 languages.
How many words is 1,000 Mistral tokens?
For typical English prose, 1,000 tokens is roughly 700–750 words. Code, JSON and non-English text use more tokens per word, so count the actual text when budgeting.
How much does Mistral Large 3 cost per token?
Mistral Large 3 costs $0.5 per 1M input tokens and $1.5 per 1M output tokens. That's $0.00050 per 1,000 input tokens.