Mistral AI
Mistral: Mistral Nemo
mistralai/mistral-nemo
A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.
- Context window: 131,072 tokens
- Input: text
- Output: text
- Pricing: $0.019/M input tokens, $0.03/M output tokens
Community sentiment
Mixed (+0) — 1 positive, 1 negative, 1 neutral across 3 community samples from X and the web, as of 2026-10-06.
Community comments praise Mistral Nemo for reliability and cost-efficiency in production/enterprise use, while some users report it loses adherence to system instructions on longer inputs. Pricing and token-throughput are noted positively in comparisons and platform listings.
- “@MistralAI Great job. You guys have some very solid and reliable models. Still rocking your 2024 mistral-nemo and it still holds up great for enterprise use case 👌” — X
- “@MistralAI I want a small model, like an improved Mistral Nemo. It's not a bad model but the longer the text you feed, less and less it follows the system instructions. Gemma has some good models but much more expensive than Mistral Nemo” — X
- “What $1 buys on OmniRail today, live prices: · 102k output tokens of Claude Sonnet 5.5, or 847k of GPT-5.6 Luna, or 34.5M of Mistral Nemo · 66 FLUX.2 Klein images, or 15 from GPT Image 2 · 43 seconds of Seedance video, 20 of Veo 3.1 Lite · 1.7M characters read aloud by Kokoro · 8” — X
View on OpenRouter.
Model data sourced from OpenRouter.