DeepSeek-V4-Flash-Vision-Exp

Fiche modèle · DeepSeek · deepseek-v4-flash-vision-exp

DeepSeek-V4-Flash-Vision-Exp is an experimental multimodal vision-understanding model on the DeepSeek API (`model='deepseek-v4-flash-vision-exp'`). It accepts text and image inputs and produces text. DeepSeek reports pure-text agent, reasoning, and world-knowledge performance on par with official DeepSeek-V4-Flash, with a large jump on agent benchmarks that require visual understanding. Thinking mode is default and supports low / high / max effort; the model supports JSON output, tool calls, the Responses API, and the Anthropic API. FIM completion is not supported. Context length is 1M tokens with up to 384K max output. Same peak/off-peak token pricing as V4-Flash; images are converted to input tokens (upper bound 384 tokens per image after resize).

Identité

OrganisationDeepSeek
Pays—
Famille—
Typellm
Date de sortie2026-08-21
Licenceunknown
Licence commercialenon
Licence open sourcenon
Poids ouvertsnon
Auto-hébergement (DGX Spark)non — poids fermés
Modalitésimage|text
Fenêtre de contexte1 000 000
KV cache / token (Q8)—
Paramètres totaux—
Paramètres actifs (session)13 B
Source des paramètres actifsweb : https://arxiv.org/abs/2606.19348
Curation des paramètres actifs2026-10-19
Prix par session0.70 $ (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 Players69.4×
Coût / tâche (Terminal-Bench)5.91 $ (est.)
Tokens d'entraînement—
Coupure de connaissance—
Créé le2026-08-21T17:51:18.063993Z
Mis à jour le2026-11-04T06:04:40.873364Z
SourceLLM Stats
URLhttps://llm-stats.com/models/deepseek-v4-flash-vision-exp

Prix par fournisseur

Statut du modèle (offres + auto-hébergement) : non.

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 « — ».

FournisseurSourceSouverainetéEntrée (/M tok)Sortie (/M tok)QuantificationContexteLatence (s)Débit (tok/s)Statut
DeepInfrallm-statsnon0.44 $1.32 $————active

Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.

Souveraineté

Scores par catégorie (top_scores)

CatégorieScore
multimodal0.30
reasoning0.40
general0.50
agents0.50
code0.70
tool_calling0.50
vision0.30

Scores par benchmark

BenchmarkCatégorieScoreMaxNormalisé (API)VérifiéAuto-déclaréDaté du
agents-last-examagents0.271.000.27nonoui2026-11-04T06:04:40.933578Z
automationbenchagents0.261.000.26nonoui2026-11-04T06:04:40.923139Z
deepswecode0.591.000.59nonoui2026-11-04T06:04:40.903378Z
dsbench-hardcode0.641.000.64nonoui2026-11-04T06:04:40.913364Z
nl2repocode0.581.000.58nonoui2026-11-04T06:04:40.893093Z
terminal-bench-2.1reasoning0.841.000.84nonoui2026-11-04T06:04:40.883880Z
zerobenchvision0.35100.000.35nonoui2026-11-04T06:04:40.943054Z

Classements

CatégorieRangScoreNote conservatriceBenchmarks évaluésMéthodeDaté duProvenance
agents54—29.836trueskill2026-11-03model_history
code52—33.084trueskill2026-11-03model_history

Lecture : Note conservatrice = borne basse TrueSkill (niveau estimé − marge d’incertitude) ; Benchmarks évalués = nombre de benchmarks de la catégorie pris en compte. À performances égales, un modèle évalué sur moins de benchmarks obtient une note plus basse : la note ne se déduit pas d’un seul score. Provenance : rankings = top 50 de /rankings ; model_history = historique du modèle (/models/{id}/history), utilisé pour les rangs au-delà du top 50.

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