Railwail
Prijava

Codestral

KodNije dostupno
od Mistral AIID modela: codestral

Mistral's code-specialized model. Optimized for code generation, completion, and understanding across 80+ languages.

Pogledaj alternativeUsporedi
Status
Nije dostupno
Kontekst
256.000 tokena
Maks. izlaz
8.192 tokena
Ulaz → izlaz
Tekst → Tekst
Razvojni inženjer
Mistral AI
Ažurirano
23. septembar 2026.

Codestral trenutno nije dostupan

Možeš i dalje pročitati detalje na ovoj stranici. Odaberi jednu od dostupnih alternativa ispod kako bi odmah pokrenuo usporediv model.

Idi na alternative
01

Usporedivi modeli

Svi u ovoj kategoriji
02

Playground

Isprobaj Codestral

Razgovor

Trenutno nije dostupno

Trenutno nedostupno.

Playground je onemogućen. Usporedive modele možete pronaći u istoj kategoriji: Pogledaj alternative

Isprobaj Codestral

Pošalji poruku. Odgovor stigne u cijelosti kada je model gotov (bez streaminga).

Sistemski prompt
Maks. duljina odgovora (tokeni)

Ovaj pokus

Nema cijene – trenutno nije dostupno.

Novi si ovdje?

10 besplatnih kredita (0,10 US$) kada se registriraš s Googlom

Dostupno 24 sati nakon registracije, do 5 pokusa dnevno i najviše 2 kredita po pokusu. Ostale metode prijave počinju bez kredita.

03

O Codestral

UkratkoStanje: 23. septembar 2026.

Codestral je model od Mistral AI iz kategorije Kod. Codestral trenutno nije dostupan na Railwailu. Kontekstni prozor sadrži 256.000 tokena, a jedan odgovor može biti dugačak do 8.192 tokena.

Pozadina

O Mistral AI

Osnovano 2023 · Paris, France

Mistral AI was founded in April 2023 in Paris by Arthur Mensch (CEO, former DeepMind), Guillaume Lample and Timothée Lacroix (both former Meta FAIR co-authors of the LLaMA papers). Mistral has built a frontier-scale European LLM lab with a portfolio mixing fully open-weight releases (Mistral 7B, Mixtral 8x7B, Mixtral 8x22B, Mistral Small) and commercial closed-API models (Mistral Large, Mistral Embed, Mistral Saba). The company has raised over €1B from investors including Andreessen Horowitz, General Catalyst, Lightspeed, Salesforce, Nvidia and Microsoft, with a 2024 valuation around €6B. Codestral was released May 2024 as Mistral's first dedicated code model — a 22B dense transformer trained on 80+ programming languages with native fill-in-the-middle support for IDE integrations. A successor, Codestral 25.01, followed in January 2025.

Posjetite Mistral AI

Arhitektura

Decoder-only Transformer for code (Mistral architecture)

Codestral-22B is a 22B dense decoder-only transformer using Mistral's standard architecture: 56 layers, 6,144 hidden size, 48-head grouped-query attention with 8 KV heads, sliding-window attention (4,096-token window), RoPE positional embeddings with theta=1M (for long-context support), SwiGLU activations and RMSNorm. The tokeniser is the Mistral BPE with 32,768 entries plus added fill-in-the-middle special tokens (`[PREFIX]`, `[SUFFIX]`, `[MIDDLE]`). Training combined a standard left-to-right autoregressive objective with FIM training for IDE-grade code completion. The training corpus covers 80+ programming languages drawn from public code repositories, plus natural-language code documentation pairs and instruction data for the chat / explain capability. Released May 2024 under the Mistral AI Non-Production License (MNPL), which permits open-weights research and evaluation but not commercial production use without a separate license or hosted API access.

Parametri
22B (dense)
Kontekst
32.000 tokena

Mogućnosti

  • 22B dense transformer purpose-built for code
  • Native fill-in-the-middle (FIM) for IDE autocomplete
  • Trained on 80+ programming languages
  • 32K context window
  • Competitive with CodeLlama-34B and DeepSeek-Coder-33B at smaller size
  • Strong on mainstream: Python, JS/TS, Java, C/C++, Go, Rust, Bash, SQL, PHP, Swift
  • Open weights available under MNPL (research only)
  • Best for: IDE code completion, code generation, explanation and refactoring across mainstream languages.

Obuka i licenca

Trained on trillions of tokens (exact figure not disclosed) of public code repositories across 80+ programming languages, natural-language code-documentation pairs, and instruction-style code data for chat and explain capabilities. Knowledge cutoff approximately early 2024. Training combined autoregressive and fill-in-the-middle objectives.

Licenca: Mistral AI Non-Production License (MNPL). Open weights for research, evaluation and personal use; commercial production requires Mistral commercial license or hosted API access (la Plateforme, AWS Bedrock, Azure AI Studio).

Testiranje sigurnosti: Mistral publishes a model card for Codestral. Safety focus is on code-specific risks (malware, vulnerability generation) rather than general consumer-chat alignment.

Poznata ograničenja

  • MNPL blocks commercial production use of open weights
  • Smaller than successor Codestral 25.01 and below DeepSeek-Coder V2 on top benchmarks
  • 32K context shorter than newer code models (DeepSeek-Coder V2: 128K)
  • Weaker on rare languages and DSLs
  • Not as strong on agentic / repo-level reasoning as longer-context code models
  • No multimodal input (no image-of-code understanding)

Znanstveni članci

04

Cijene

Trenutno nedostupno. Za ovaj model trenutno nema cijene, pa se ne može pokrenuti.

05

API

Pozovi Codestral sa svojim Railwail API ključem. Koristi ovaj ID modela u zahtjevu:
codestralPogledaj API dokumentacijuDobij API ključ

Trenutno nije dostupno

Model nema provjeren cijenu ili je deaktiviran; API pozivi se odbijaju.

06

Specifikacije

ID modela
codestral
Razvojni inženjer
Mistral AI
Kategorija
Kod
Ulaz
Tekst
Izlaz
Tekst
Kontekstni prozor
256.000 tokena
Maks. izlaz
8.192 tokena
Veličina modela
22B (dense)
Licenca
Mistral AI Non-Production License (MNPL). Open weights for research, evaluation and personal use; commercial production requires Mistral commercial license or hosted API access (la Plateforme, AWS Bedrock, Azure AI Studio).
Katalog ažuriran
23. septembar 2026.

Oznake

  • coding
  • fast
  • multilanguage
07

Primjer promptova

Primjeri iz Railwail kataloga. Nisu generirani uživo na ovoj stranici.

  • TypeScript API Handler

    Write a type-safe Express middleware for JWT authentication in TypeScript
    Prikaži primjer odgovora

    import { Request, Response, NextFunction } from 'express'; import jwt from 'jsonwebtoken'; interface JwtPayload { userId: string; email: string; role: 'admin' | 'user'; } declare global { namespace Express { interface Request { user?: JwtPayload; } } } export const authenticate = ( req: Request, res: Response, next: NextFunction ): void => { const token = req.headers.authorization?.split(' ')[1]; if (!token) { res.status(401).json({ error: 'No token provided' }); return; } try { const decoded = jwt.verify( token, process.env.JWT_SECRET! ) as JwtPayload; req.user = decoded; next(); } catch (err) { res.status(403).json({ error: 'Invalid or expired token' }); } };

  • Rust Error Handling

    Write a Rust function that reads a JSON config file and returns a parsed struct with proper error handling
    Prikaži primjer odgovora

    use serde::Deserialize; use std::fs; use std::path::Path; use thiserror::Error; #[derive(Debug, Deserialize)] pub struct AppConfig { pub host: String, pub port: u16, pub database_url: String, pub max_connections: Option<u32>, } #[derive(Error, Debug)] pub enum ConfigError { #[error("Failed to read config file: {0}")] IoError(#[from] std::io::Error), #[error("Failed to parse config JSON: {0}")] ParseError(#[from] serde_json::Error), #[error("Config file not found: {0}")] NotFound(String), } pub fn load_config(path: &str) -> Result<AppConfig, ConfigError> { if !Path::new(path).exists() { return Err(ConfigError::NotFound(path.to_string())); } let contents = fs::read_to_string(path)?; let config: AppConfig = serde_json::from_str(&contents)?; Ok(config) } // Usage: // let config = load_config("config.json")?;

08

Primjena

Za što se koristi

  • IDE code completion (continue.dev, JetBrains)
  • Code generation and refactoring
  • Code explanation and review
  • Cross-language code translation
  • Self-hosted developer assistants (research)
  • Production deployments via Mistral hosted API
09

Često postavljana pitanja

Što je Codestral?

Codestral je model od Mistral AI iz kategorije Kod. Naveden je na Railwailu, ali se trenutno ne može pokrenuti.

Koliko stoji Codestral na Railwailu?

Codestral se trenutno ne može pokrenuti na Railwailu, pa nema trenutne cijene. Dostupne alternative s cijenama navedene su dalje na ovoj stranici.

Koliki je kontekstni prozor od Codestral?

Kontekstni prozor od Codestral sadrži 256.000 tokena. Jedan odgovor može biti dugačak do 8.192 tokena.

Koliko je brz Codestral?

Za Codestral na Railwailu još nema dovoljno izmjerenih pokretanja da bi se navedlo vrijeme izvršavanja. Ovisi o unošenju, postavkama i opterećenju kod pružatelja.

Je li Codestral bolji od Code Llama 13B Instruct?

To ovisi o zadatku. Codestral (Mistral AI) i Code Llama 13B Instruct (Meta) su oba modela iz kategorije Kod. Stranica za usporedbu prikazuje njihove cijene i specifikacije usporedno.

Usporedi Codestral i Code Llama 13B Instruct

Mogu li koristiti Codestral sada?

Trenutno nedostupno. Stranica ostaje dostupna; dostupne alternative iz iste kategorije navedene su dalje.

Svi modeli preko jednog API-ja

Jedan API ključ za sve modele na Railwailu. Korištenje se naplaćuje iz unaprijed plaćenih kredita, 1 kredit = 0,01 US$.

Dobij API ključPogledaj sve modele