KI‑Modell‑Intelligenz

Anbieter · 2026-05-12

NVIDIA

37 kanonische Modelle51 Einträge insgesamt (inkl. Derivate)
ModellEingabe / 1MAusgabe / 1MKontextHosterTags
nvidia-nemotron-nano-9b-v2$0.040$0.160131K6tools · reasoning · open-weights
nemotron-3-nano-30b-a3b$0.050$0.200131K6tools · reasoning · open-weights
Llama 3.3 Nemotron Super 49B v1.5$0.050$0.250131K3tools · reasoning · open-weights
Llama 3.3 Nemotron Super 49B v1$0.150$0.150131K2tools · reasoning · open-weights
Nemotron 3 Super$0.200$0.800262K8tools · reasoning · open-weights
Active Speaker DetectionUnbekanntUnbekanntUnknown1open-weights
bevformerUnbekanntUnbekannt128K1open-weights
BGE M3UnbekanntUnbekannt8K1open-weights
cosmos-predict1-5bUnbekanntUnbekanntUnknown1vision · open-weights
cosmos-transfer1-7bUnbekanntUnbekanntUnknown1vision · open-weights
cosmos-transfer2.5-2bUnbekanntUnbekanntUnknown1vision · open-weights
FLUX.1-devUnbekanntUnbekannt4K1
FLUX.1-Kontext-devUnbekanntUnbekannt41K1vision · open-weights
FLUX.1-schnellUnbekanntUnbekannt771open-weights
FLUX.2 Klein 4BUnbekanntUnbekannt41K1vision · open-weights
gliner-piiUnbekanntUnbekannt128K1open-weights
llama-3_2-nemoretriever-300m-embed-v1UnbekanntUnbekannt33K1open-weights
llama-3.1-nemotron-safety-guard-8b-v3UnbekanntUnbekannt128K1open-weights
llama-nemotron-embed-vl-1b-v2UnbekanntUnbekannt33K1vision · open-weights
llama-nemotron-rerank-vl-1b-v2UnbekanntUnbekannt128K1vision · open-weights
magpie-tts-zeroshotUnbekanntUnbekanntUnknown1open-weights
Nemotron 3 Nano OmniUnbekanntUnbekannt256K1tools · json · reasoning · vision · open-weights
nemotron-3-content-safetyUnbekanntUnbekannt128K1open-weights
nemotron-content-safety-reasoning-4bUnbekanntUnbekannt128K1reasoning · open-weights
nemotron-mini-4b-instructUnbekanntUnbekannt128K1tools · open-weights
nemotron-voicechatUnbekanntUnbekannt128K1tools · open-weights
nv-embed-v1UnbekanntUnbekannt33K1open-weights
nv-embedcode-7b-v1UnbekanntUnbekannt33K1open-weights
rerank-qa-mistral-4bUnbekanntUnbekannt128K1open-weights
riva-translate-4b-instruct-v1_1UnbekanntUnbekannt128K1open-weights
sarvam-mUnbekanntUnbekannt128K1tools · open-weights
sparsedriveUnbekanntUnbekannt128K1open-weights
streampetrUnbekanntUnbekannt128K1open-weights
studiovoiceUnbekanntUnbekannt128K1open-weights
synthetic-video-detectorUnbekanntUnbekanntUnknown1open-weights
usdcodeUnbekanntUnbekannt128K1
usdvalidateUnbekanntUnbekanntUnknown1open-weights
ByteDance-Seed/Seed-OSS-36B-InstructDerivatUnbekanntUnbekannt262K1tools · json
Gemma 2 2b ItDerivatUnbekanntUnbekannt128K1tools · json · open-weights
GLM-4.7DerivatUnbekanntUnbekannt205K1tools · reasoning · open-weights
Magistral Small 2506DerivatUnbekanntUnbekannt33K1
Mistral Large 3 675B Instruct 2512DerivatUnbekanntUnbekannt262K1tools · json · vision · open-weights
Mistral Medium 3DerivatUnbekanntUnbekannt131K1vision
Mistral-7B-Instruct-v0.3DerivatUnbekanntUnbekannt66K1tools · json · open-weights
mistral-nemotronDerivatUnbekanntUnbekannt128K1tools · open-weights
mistral-small-4-119b-2603DerivatUnbekanntUnbekannt128K1tools · open-weights
Mistral: Mixtral 8x7B InstructDerivatUnbekanntUnbekannt33K1tools · open-weights
paligemmaDerivatUnbekanntUnbekannt128K1vision · open-weights
Qwen Image EditDerivatUnbekanntUnbekanntUnknown1vision
solar-10.7b-instructDerivatUnbekanntUnbekannt128K1tools · open-weights
Whisper Large v3DerivatUnbekanntUnbekanntUnknown1open-weights

Frequently asked questions

How many AI models does NVIDIA offer?

We track 37 canonical NVIDIA models plus 14 community fine-tunes / derivatives (excluded from the main table). The list is recomputed daily from models.dev.

Which NVIDIA model is the cheapest?

nvidia-nemotron-nano-9b-v2 is currently the lowest-priced NVIDIA model, at $0.040 per 1M input tokens and $0.160 per 1M output tokens. For the full apples-to-apples list, see /pricing/cheapest-llm-api.

Which NVIDIA model has the largest context window?

Nemotron 3 Super leads at 262K tokens. This is the total of prompt + completion.

Which NVIDIA models support tool calling?

Multiple NVIDIA models support tool calling, with nvidia-nemotron-nano-9b-v2 being a popular pick. The capability column in the table above marks every model with NVIDIA tool-calling support.

Which NVIDIA models accept image input?

cosmos-predict1-5b accepts image input. Other vision-capable NVIDIA models are tagged 'vision' in the table above. See /capabilities/vision for a cross-vendor comparison.

What are the best alternatives to NVIDIA?

Depends on the use case. For raw cost savings, look at /pricing/cheapest-llm-api. For agent-oriented workloads, /best/best-ai-model-for-agents. For long-document workflows, /best/best-long-context-llm.

How fresh is this NVIDIA pricing data?

Daily. Our pipeline pulls models.dev each morning and rebuilds these pages on data change, so list-price moves and new model releases land within roughly 24 hours.

Zuletzt aktualisiert:

Prices in USD per 1M tokens. Unknown means the provider does not publish per-token pricing.

Data is sourced from models.dev and normalized for comparison. Prices and capabilities may change. Always verify critical production decisions with the provider's official documentation.