IBM Granite 4.2 3B
IBM Granite 4.2 3B is the compact dense decoder-only reasoning LLM in the Granite 4.2 family. It supports native think, non-think, and low-effort modes, tool calling, and a 128K native context window with extension to 512K. Unlike the 8B and 30B, it did not receive the extra agentic RL block. Open weights on Hugging Face under Apache 2.0. Text-only (not multimodal).
Identité
| Organisation | IBM |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | 2026-08-25 |
| Licence | apache 2.0 |
| Licence commerciale | oui |
| Licence open source | oui |
| Poids ouverts | oui |
| Auto-hébergement (DGX Spark) | oui — ≤ 107 Md (Q8/FP8) |
| Modalités | text |
| Fenêtre de contexte | 131 072 |
| KV cache / token (Q8) | 41.0 ko/token |
| Paramètres totaux | 3 B |
| Paramètres actifs (session) | 3.66 B |
| Source des paramètres actifs | actifs = total (modèle dense, config HF (ibm-granite/granite-4.2-3b)) ; total : config Hugging Face (ibm-granite/granite-4.2-3b) |
| Curation des paramètres actifs | 2026-11-04 |
| Prix par session | 0.05 $ (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) | 0.45 $ (est.) |
| Tokens d'entraînement | — |
| Coupure de connaissance | — |
| Créé le | 2026-08-30T23:49:06.642023Z |
| Mis à jour le | 2026-11-04T06:04:47.833274Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/granite-4.2-3b |
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 | 0.03 $ | 0.12 $ | — | — | — | — | active |
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| instruction_following | 0.70 |
| knowledge | 0.30 |
| language | 0.70 |
| legal | 0.70 |
| long_context | 0.60 |
| math | 0.60 |
| physics | 0.40 |
| reasoning | 0.50 |
| finance | 0.70 |
| general | 0.60 |
| healthcare | 0.70 |
| agents | 0.40 |
| biology | 0.40 |
| chemistry | 0.40 |
| code | 0.20 |
| creativity | 0.30 |
| tool_calling | 0.50 |
| writing | 0.30 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| aime-2025 | math | 0.78 | 1.00 | 0.78 | non | oui | 2026-11-04T06:04:47.863128Z |
| arena-hard-v2 | general | 0.35 | 1.00 | 0.35 | non | oui | 2026-11-04T06:04:47.913005Z |
| bfcl-v4 | tool_calling | 0.52 | 1.00 | 0.52 | non | oui | 2026-11-04T06:04:47.843045Z |
| gpqa | reasoning | 0.55 | 1.00 | 0.55 | non | oui | 2026-11-04T06:04:47.883037Z |
| hmmt25 | math | 0.67 | 1.00 | 0.67 | non | oui | 2026-11-04T06:04:47.873103Z |
| ifbench | instruction_following | 0.74 | 1.00 | 0.74 | non | oui | 2026-11-04T06:04:47.921173Z |
| livecodebench-v6 | reasoning | 0.70 | 1.00 | 0.70 | non | oui | 2026-11-04T06:04:47.887614Z |
| mmlu-pro | language | 0.68 | 1.00 | 0.68 | non | oui | 2026-11-04T06:04:47.903071Z |
| profbench | reasoning | 0.32 | 1.00 | 0.32 | non | oui | 2026-11-04T06:04:47.853101Z |
| ruler-128k | long_context | 0.55 | 1.00 | 0.55 | non | oui | 2026-11-04T06:04:47.937582Z |
| ruler-64k | long_context | 0.68 | 1.00 | 0.68 | non | oui | 2026-11-04T06:04:47.933209Z |
| scicode | reasoning | 0.24 | 1.00 | 0.24 | non | oui | 2026-11-04T06:04:47.892927Z |
| tau3-bench | agents | 0.46 | 1.00 | 0.46 | non | oui | 2026-11-04T06:04:47.837685Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| knowledge | 34 | 0.30 | -1.61 | 1 | trueskill | 2026-11-04T08:08:45.139626Z | rankings |
| agents | 186 | — | 1.10 | 3 | trueskill | 2026-10-19 | model_history |
| code | 277 | — | -7.18 | 1 | 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