Gemma 3 4B
Gemma 3 4B is a 4-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for question answering, summarization, reasoning, and image understanding tasks.
Identité
| Organisation | |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | 2025-03-12 |
| Licence | gemma |
| Licence commerciale | oui |
| Licence open source | non |
| Poids ouverts | oui |
| Auto-hébergement (DGX Spark) | oui — ≤ 107 Md (Q8/FP8) |
| Modalités | image|text |
| Fenêtre de contexte | 128 000 |
| KV cache / token (Q8) | — |
| Paramètres totaux | 4 B |
| Paramètres actifs (session) | 4.3 B |
| Source des paramètres actifs | web : https://huggingface.co/google/gemma-3-4b-it |
| Curation des paramètres actifs | 2026-11-04 |
| Prix par session | 0.07 $ (1 000 000 tokens entrée + 200 000 sortie, prix le moins cher par fournisseur) |
| Tokens / tâche (Data Players) | 1.2 M |
| Tokens / tâche (Terminal-Bench) | 83.3 M (est.) |
| Ratio Terminal-Bench / Data Players | 69.4× |
| Coût / tâche (Terminal-Bench) | 0.59 $ (est.) |
| Tokens d'entraînement | 4 000 000 000 000 |
| Coupure de connaissance | 2024-08-01 |
| Créé le | 2026-05-07T16:53:29.688684Z |
| Mis à jour le | 2026-11-04T06:04:45.586947Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/gemma-3-4b-it |
Prix par fournisseur
Statut du modèle (offres + auto-hébergement) : partiel — auto-hébergement DGX Spark (≤ 107 Md).
Toutes sources confondues : catalogue llm-stats, routeurs EUrouter/Cortecs et catalogues directs (OVHcloud, Scaleway, Infomaniak). Les prix sont convertis en USD pour la comparaison ; le prix publié et le taux BCE utilisé sont dans l'infobulle. Les prix inconnus restent indiqués « — ».
| Fournisseur | Source | Souveraineté | Entrée (/M tok) | Sortie (/M tok) | Quantification | Contexte | Latence (s) | Débit (tok/s) | Statut |
|---|---|---|---|---|---|---|---|---|---|
| DeepInfra | llm-stats | non | 0.05 $ | 0.10 $ | — | — | — | — | active |
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| chat | 0.90 |
| image_to_text | 0.70 |
| instruction_following | 0.90 |
| language | 0.40 |
| legal | 0.40 |
| math | 0.60 |
| multimodal | 0.60 |
| physics | 0.30 |
| reasoning | 0.50 |
| structured_output | 0.90 |
| factuality | 0.40 |
| finance | 0.40 |
| general | 0.50 |
| grounding | 0.70 |
| healthcare | 0.50 |
| biology | 0.30 |
| chemistry | 0.30 |
| code | 0.40 |
| vision | 0.60 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| ai2d | vision | 0.75 | 1.00 | 0.75 | non | oui | 2026-11-04T06:04:45.593489Z |
| big-bench-extra-hard | reasoning | 0.11 | 1.00 | 0.11 | non | oui | 2026-11-04T06:04:45.603153Z |
| big-bench-hard | reasoning | 0.72 | 1.00 | 0.72 | non | oui | 2026-11-04T06:04:45.608004Z |
| bird-sql-(dev) | reasoning | 0.36 | 1.00 | 0.36 | non | oui | 2026-11-04T06:04:45.613095Z |
| chartqa | reasoning | 0.69 | 1.00 | 0.69 | non | oui | 2026-11-04T06:04:45.618636Z |
| docvqa | vision | 0.76 | 1.00 | 0.76 | non | oui | 2026-11-04T06:04:45.623047Z |
| eclektic | language | 0.05 | 1.00 | 0.05 | non | oui | 2026-11-04T06:04:45.628684Z |
| facts-grounding | reasoning | 0.70 | 1.00 | 0.70 | non | oui | 2026-11-04T06:04:45.632583Z |
| global-mmlu-lite | general | 0.55 | 1.00 | 0.55 | non | oui | 2026-11-04T06:04:45.643069Z |
| gpqa | reasoning | 0.31 | 1.00 | 0.31 | non | oui | 2026-11-04T06:04:45.646152Z |
| gsm8k | math | 0.89 | 1.00 | 0.89 | non | oui | 2026-11-04T06:04:45.653010Z |
| hiddenmath | math | 0.43 | 1.00 | 0.43 | non | oui | 2026-11-04T06:04:45.656137Z |
| humaneval | reasoning | 0.71 | 1.00 | 0.71 | non | oui | 2026-11-04T06:04:45.659092Z |
| ifeval | chat | 0.90 | 1.00 | 0.90 | non | oui | 2026-11-04T06:04:45.662910Z |
| infovqa | vision | 0.50 | 1.00 | 0.50 | non | oui | 2026-11-04T06:04:45.669213Z |
| livecodebench | reasoning | 0.13 | 1.00 | 0.13 | non | oui | 2026-11-04T06:04:45.673172Z |
| math | math | 0.76 | 1.00 | 0.76 | non | oui | 2026-11-04T06:04:45.679972Z |
| mathvista-mini | math | 0.50 | 1.00 | 0.50 | non | oui | 2026-11-04T06:04:45.688290Z |
| mbpp | reasoning | 0.63 | 100.00 | 0.63 | non | oui | 2026-11-04T06:04:45.693189Z |
| mmlu-pro | language | 0.44 | 1.00 | 0.44 | non | oui | 2026-11-04T06:04:45.697283Z |
| mmmu-(val) | vision | 0.49 | 1.00 | 0.49 | non | oui | 2026-11-04T06:04:45.703106Z |
| natural2code | reasoning | 0.70 | 1.00 | 0.70 | non | oui | 2026-11-04T06:04:45.707393Z |
| simpleqa | factuality | 0.04 | 1.00 | 0.04 | non | oui | 2026-11-04T06:04:45.713086Z |
| textvqa | vision | 0.58 | 1.00 | 0.58 | non | oui | 2026-11-04T06:04:45.716442Z |
| vqav2-(val) | vision | 0.62 | 1.00 | 0.62 | non | oui | 2026-11-04T06:04:45.727385Z |
| wmt24++ | language | 0.47 | 1.00 | 0.47 | non | oui | 2026-11-04T06:04:45.733138Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| chat | 49 | 0.90 | 20.21 | 1 | trueskill | 2026-11-04T08:08:43.789169Z | rankings |
| code | 270 | — | -4.41 | 2 | trueskill | 2026-10-19 | model_history |
Données
Dates de curation (récupération depuis l'API) : models 2026-11-04 · benchmarks 2026-11-04 · scores 2026-11-04 · updates 2026-11-04 · account 2026-11-04 · model_history 2026-11-04 · rankings 2026-11-04