Mistral NeMo 12B — Spezifikationen
Compiled by Mustafa Ihsan from the vendor’s published documentation · Last updated
| Entwickler | Mistral AI |
|---|---|
| Typ | LLM (dicht) |
| Modality | Text → Text |
| Parameter | 12B |
| Kontextfenster | 128 K |
| Lizenz | Apache 2.0 (offen) |
| Offene Gewichte | Ja |
| Veröffentlicht | 2024 |
| Eingabepreis | $0.02 /1M |
| Ausgabepreis | $0.04 /1M |
| API-Anbieter | Mistral, OpenRouter, Ollama |
Lokal ausführen
| VRAM (4-Bit) | ~7,5 GB |
|---|---|
| Mindest-GPU | RTX 4070 12 GB / RTX 3060 |
What is Mistral NeMo 12B?
Mistral NeMo 12B was built by Mistral AI in partnership with NVIDIA — Apache 2.0, a 128K
context, strong multilingual support, and about 7.5 GB of VRAM at 4-bit, which fits an RTX
4070 12GB or an RTX 3060.
The NVIDIA collaboration shows up where it matters: the model was designed with quantisation
in mind, so it holds quality at 4-bit better than models retrofitted to small hardware after
the fact — the usual failure mode being a model that technically fits in 12 GB but degrades
noticeably once it does. Combined with genuinely good multilingual coverage, that makes NeMo
12B a strong pick for non-English local deployments, where the alternatives are often
English-first models that lose more in quantisation than they can afford. At $0.02 in / $0.04
out per million tokens it is also among the cheapest hosted options available, so it works as
the low-cost tier of a routing setup as well as a local model.
Mistral NeMo 12B pricing: API cost per 1M tokens
| Eingabe (pro 1 Mio. Token) | $0.0200 |
|---|---|
| Ausgabe (pro 1 Mio. Token) | $0.0400 |
| Verhältnis Output/Input | 2× |
| Gemischt (4:1 Input:Output) | $0.0240 pro 1 Mio. Tokens |
What Mistral NeMo 12B costs per month
Tatsächliche monatliche Ausgaben bei einem 4:1-Input-zu-Output-Mix – dem Verhältnis, das typische Chat- oder RAG-Arbeitslasten tatsächlich erzeugen.
| Workload | Tokens/Monat | Kosten pro Monat |
|---|---|---|
| Nebenprojekt | 1 Mio. Eingabe / 0,25 Mio. Ausgabe | $0.03 |
| Kleines Team | 20 Mio. Eingabe / 5 Mio. Ausgabe | $0.60 |
| Produktion | 200 Mio. Eingabe / 50 Mio. Ausgabe | $6.00 |
Stellen Sie Ihre eigenen Berechnungen im KI-API-Kostenrechner.
Cheaper alternatives to Mistral NeMo 12B
| Modell | Gewichteter Preis pro Million US-Dollar | Sie sparen |
|---|---|---|
| Mistral 7B offenere | $0.0220 | 8% cheaper |
| Llama 3.1 8B offenere | $0.0220 | 8% cheaper |
Selbst hosten oder die API nutzen?
Mistral NeMo 12B is open-weight, so you can run it yourself. It needs ~7,5 GB of VRAM at 4-bit (RTX 4070 12GB / RTX 3060). Self-hosting only beats the API once your volume is high enough to keep that hardware busy — the Selbsthosting vs. API-Rechner berechnet den Break-even-Punkt für Ihr Token-Volumen.
Häufig gestellte Fragen
How much does Mistral NeMo 12B cost per 1M tokens?
Mistral NeMo 12B costs $0.0200 per 1M input tokens and $0.0400 per 1M output tokens. At a typical 4:1 input-to-output mix that blends to about $0.0240 per 1M tokens.
How much does Mistral NeMo 12B cost per month?
A small-team workload of 20M input and 5M output tokens a month costs about $0.60 on Mistral NeMo 12B. A side project (1M in / 0.25M out) costs roughly $0.03.
What is a cheaper alternative to Mistral NeMo 12B?
Mistral 7B is the strongest cheaper option in our database at $0.0220 per 1M blended — about 8% less than Mistral NeMo 12B. It is also open-weight, so self-hosting is an option.
Can I run Mistral NeMo 12B locally?
Yes. Mistral NeMo 12B is open-weight and needs about ~7.5 GB of VRAM at 4-bit quantisation (RTX 4070 12GB / RTX 3060).
Why does Mistral NeMo 12B charge more for output than input?
Output tokens are generated one at a time and cannot be batched the way a prompt can, so they cost the provider more to serve. Mistral NeMo 12B charges 2× more for output, which is why prompt-heavy workloads are far cheaper to run than generation-heavy ones.
Die Preise entsprechen den offiziell veröffentlichten Listenpreisen für die primäre API des jeweiligen Modells und werden regelmäßig aktualisiert, sobald Anbieter diese ändern. Volumen-, Batch- und Cached-Input-Rabatte sind nicht enthalten. Vergleichen Sie alle Modelle nebeneinander im Datenbank für KI-Modelle oder das LLM-Leaderboard.