Processing 10,000 lines of code (~98,000 tokens) with Mistral Small 4 costs $0.0147 for codebase ingestion and $0.0265 for an AI-powered code review and refactoring pass.
Quick answer
A 10,000 lines of code codebase is estimated at 98,000 tokens for Mistral Small 4. Ingestion costs $0.0147; a review and refactoring pass with roughly 20% output costs about $0.0265.
Method & trust
Code is estimated at a tokenizer-specific tokens-per-line ratio, then priced separately for repository context and generated review output. Comments, minified files, tests, and repeated context can materially change the bill.
Feed 10,000 lines of code into the prompt context for repository search, Q&A, or architecture planning.
Ingest 10,000 lines of code and generate audit findings, unit test recommendations, and refactor diffs.
Mistral Small 4 writes 10,000 lines of code from scratch based on product specifications.
| Model | Provider | Code Ingestion | Cached Ingestion | Code Review Cost | Context Limit |
|---|---|---|---|---|---|
| Mistral Small 4 (Current) | mistral | $0.0147 | $0.00147 | $0.0265 | 128,000 |
| GPT-5.6 Luna | openai | $0.0204 | $0.00204 | $0.0449 | 1,050,000 |
| GPT-5.4 mini | openai | $0.0765 | $0.00765 | $0.1683 | 256,000 |
| GPT-5.4 nano | openai | $0.0204 | $0.00204 | $0.0459 | 128,000 |
| GPT-4o mini | openai | $0.0153 | $0.00765 | $0.0275 | 128,000 |
| Claude Haiku 4.5 | anthropic | $0.112 | $0.0112 | $0.224 | 1,000,000 |
On average, code yields approximately 9.8 tokens per line in Mistral Small 4 (mistral_bpe tokenizer). Indentation, brackets, camelCase variable names, and comments slightly increase token density compared to plain English text. 10,000 lines of code produces approximately 98,000 tokens.
Sending 10,000 lines of code as context and generating a thorough code review with recommendations costs approximately $0.0265. Utilizing prompt caching on repeat turns or static repository definitions drops this to $0.0132.
Mistral Small 4 has a context window of 128,000 tokens. 10,000 lines of code consumes 76.56% of its total available context.