Free NVIDIA NIM models
Free with NVIDIA Developer Program membership. 100+ models. Rate-limited (no daily token cap).
Free NVIDIA NIM models in this directory
- DeepSeek V4 Flash · Free tier
- GLM 5.1 · Free tier
- Kimi K2.6 · Prototyping only · rate limits not published
- DeepSeek V4 Pro · Free tier
- Arctic Embed L · Prototyping only · rate limits not published
- CodeGemma 1.1 7B · Prototyping only · rate limits not published
- CodeGemma 7B · Prototyping only · rate limits not published
- CodeLlama 70B · Prototyping only · rate limits not published
- Codestral 22B Instruct v0.1 · Prototyping only · rate limits not published
- Cosmos Reason2 8B · Prototyping only · rate limits not published
- DBRX Instruct · Prototyping only · rate limits not published
- DeepSeek Coder 6.7B Instruct · Prototyping only · rate limits not published
- DeepSeek V4.1 Flash · Prototyping only · rate limits not published
- deepseek-ai/deepseek-r1 · ~40 RPM
- DiffusionGemma 26B A4B IT · Prototyping only · rate limits not published
- Embed QA 4 · Prototyping only · rate limits not published
- Gemma 2B · Prototyping only · rate limits not published
- Gemma 3 12B IT · Prototyping only · rate limits not published
- Gemma 3 4B IT · Prototyping only · rate limits not published
- GLM 5.3 Flash · Prototyping only · rate limits not published
- google/gemma-4-31b · ~40 RPM
- Granite 3.0 3B A800M Instruct · Prototyping only · rate limits not published
- Granite 3.0 8B Instruct · Prototyping only · rate limits not published
- Granite 34B Code Instruct · Prototyping only · rate limits not published
- Granite 8B Code Instruct · Prototyping only · rate limits not published
- Jamba 1.5 Large Instruct · Prototyping only · rate limits not published
- Laguna XS 2.1 · Prototyping only · rate limits not published
- Llama 2 70B · Prototyping only · rate limits not published
- Llama 3.1 Nemotron 51B Instruct · Prototyping only · rate limits not published
- Llama 3.1 Nemotron 70B Instruct · Prototyping only · rate limits not published
- Llama 3.2 90B Vision Instruct · Prototyping only · rate limits not published
- Llama 3.2 NemoRetriever 1B VLM Embed v1 · Prototyping only · rate limits not published
- Llama 3.2 NV EmbedQA 1B v1 · Prototyping only · rate limits not published
- Llama Nemotron Embed VL 1B v2 · Prototyping only · rate limits not published
- Llama3 ChatQA 1.5 70B · Prototyping only · rate limits not published
- meta/llama-3.1-405b-instruct · ~40 RPM
- MiniMax M3.1 Flash · Prototyping only · rate limits not published
- minimax/minimax-m2.7 · ~40 RPM
- Mistral 7B Instruct v0.3 · Prototyping only · rate limits not published
- Mistral Large · Prototyping only · rate limits not published
- Mistral Large 4 1T A52B · Prototyping only · rate limits not published
- Mistral NeMo 12B Instruct · Prototyping only · rate limits not published
- Mistral NeMo Minitron 8B 8K Instruct · Prototyping only · rate limits not published
- mistralai/mistral-large-2-instruct · Prototyping only · rate limits not published
- Mixtral 8x22B v0.1 · Prototyping only · rate limits not published
- Nemotron 3 Embed 1B · Prototyping only · rate limits not published
- Nemotron 3 Ultra · Prototyping only · rate limits not published
- Nemotron 3.5 Bolt E4B · Prototyping only · rate limits not published
- Nemotron 3.5 Embed 1B · Prototyping only · rate limits not published
- Nemotron 3.5 Embed 8B · Prototyping only · rate limits not published
- Nemotron 3.5 Lightning 30B A3B · Prototyping only · rate limits not published
- Nemotron 3.5 Super VL 120B A12B · Prototyping only · rate limits not published
- Nemotron Nano 3 30B A3B · Prototyping only · rate limits not published
- Nemotron-4 340B Instruct · Prototyping only · rate limits not published
- NV EmbedQA Mistral 7B v2 · Prototyping only · rate limits not published
- nvidia/llama-3.1-nemotron-ultra-253b-v1 · Prototyping only · rate limits not published
- nvidia/nemotron-3-nano-30b-a3b · ~40 RPM
- nvidia/nemotron-3-super-120b-a12b · Prototyping only · rate limits not published
- Palmyra Creative 122B · Prototyping only · rate limits not published
- Palmyra Fin 70B 32K · Prototyping only · rate limits not published
- Palmyra Med 70B · Prototyping only · rate limits not published
- Palmyra Med 70B 32K · Prototyping only · rate limits not published
- Phi-3 Vision 128K Instruct · Prototyping only · rate limits not published
- Phi-3.5 MoE Instruct · Prototyping only · rate limits not published
- qwen/qwen2.5-72b-instruct · ~40 RPM
- RecurrentGemma 2B · Prototyping only · rate limits not published
- Riva Translate 4B Instruct · Prototyping only · rate limits not published
- Riva Translate 4B Instruct v2 · Prototyping only · rate limits not published
- SEA-LION 7B Instruct · Prototyping only · rate limits not published
- StarCoder2 15B · Prototyping only · rate limits not published
- Step 5 Preview · Prototyping only · rate limits not published
- Yi Large · Prototyping only · rate limits not published
- Zamba2 7B Instruct · Prototyping only · rate limits not published
- MiniMax M2.7 · Free tier
- Step 3.5 Flash · Free tier
- Step 3.7 Flash · Free tier
- Qwen3.5 400B VLM · Free tier
- Mistral Medium 3.5 · Free tier
- Qwen3 80B Instruct · Free tier
- Qwen3.5 122B · Free tier
- Llama 4 Maverick · Free tier
- GPT OSS 120B · Free tier
- Mistral Small 4 · Free tier
- MiniMax M3 · Free tier
- Mistral Large 675B · Free tier
- Nemotron 3 Omni · Prototyping only · rate limits not published
- Gemma 4 31B · Prototyping only · rate limits not published
- GPT OSS 20B · Prototyping only · rate limits not published
- Seed OSS 36B · Free tier
- Stockmark 100B · Free tier
- Ministral 14B · Free tier
- Llama 3.2 11B Vision · Free tier
- Phi 4 Mini · Free tier
Using this model in production?
We wire free-tier and open-weight models into production stacks: evals against your real inputs, failover, cost caps. Book a free 30-minute call.
Frequently asked
Are NVIDIA NIM models free?
Yes, within NVIDIA NIM's free tier (Prototyping only · rate limits not published). No credit card is required. This directory lists 93 of its free models individually; beyond those, the provider offers 90 more models on the same tier.
Can I use NVIDIA NIM's free models commercially?
Commercial terms are not clearly documented. Check the provider's current terms before shipping.
How do I start using NVIDIA NIM's free models?
Free with NVIDIA Developer Program membership. 100+ models. Rate-limited (no daily token cap).
Free monthly update
The monthly free-models update
What is newly free, what got rate-limited, and what to switch to. One email a month. Unsubscribe anytime.
Related free models
- GLM 4.6 (Z.ai) · Self-host free
- DeepSeek V3-0324 (DeepSeek) · Self-host free
- Gemini 2.5 Flash (Google AI Studio) · Free tier · existing users only, shuts down Dec 1, 2026
- Gemini 2.5 Pro (Google Gemini) · Free tier · existing users only, shuts down Dec 1, 2026
- Qwen3.7 Max (Qwen) · One-time 1M tokens, 90 days (new users)
- GPT OSS 120B (Scaleway) · First 1M tokens free (shared) · then paid