modelgrep

Best Free LLMs

Quick answer · Updated June 2026

The most capable free LLM is Nemotron 3 Ultra (free), scoring 47.7 on the Intelligence Index at no per-token cost via OpenRouter. Gemma 4 31B (free) (39.2) and Nemotron 3 Super (free) (36.0) round out the top three.

47.7Intelligence
64 t/sSpeed
FreeInput /M
1MContext

The best large language models with a free tier, ranked by intelligence. Capable AI models you can use at no cost via OpenRouter.

  1. 1N
    nemotron-3-ultra-550b-a55b:free
    ReasoningTools47.7 intel · 64 t/s · 1.2s ttft
    47.7
    Intelligence
  2. 2G
    gemma-4-31b-it:free
    ReasoningToolsJSON+139.2 intel · 64 t/s · 269ms ttft
    39.2
    Intelligence
  3. 3N
    nemotron-3-super-120b-a12b:free
    ReasoningToolsJSON36.0 intel · 240 t/s · 1.2s ttft
    36.0
    Intelligence
  4. 4G
    gemma-4-26b-a4b-it:free
    ReasoningToolsJSON+131.2 intel · 56 t/s · 366ms ttft
    31.2
    Intelligence
  5. 5Q
    qwen3-coder:free
    Tools24.8 intel · 45 t/s · 654ms ttft
    24.8
    Intelligence
  6. 6O
    gpt-oss-120b:free
    ReasoningTools24.5 intel · 450 t/s · 181ms ttft
    24.5
    Intelligence
  7. 7O
    gpt-oss-20b:free
    ReasoningToolsJSON24.5 intel · 348 t/s · 235ms ttft
    24.5
    Intelligence
  8. 8N
    nemotron-3-nano-30b-a3b:free
    ReasoningTools24.3 intel · 141 t/s · 434ms ttft
    24.3
    Intelligence
  9. 9N
    nemotron-3-nano-omni-30b-a3b-reasoning:free
    ReasoningToolsVision+121.4 intel · 194 t/s · 436ms ttft
    21.4
    Intelligence
  10. 10Q
    qwen3-next-80b-a3b-instruct:free
    ToolsJSON20.1 intel · 81 t/s · 476ms ttft
    20.1
    Intelligence
  11. 11N
    hermes-3-llama-3.1-405b:free
    17.6 intel · 23 t/s · 339ms ttft
    17.6
    Intelligence
  12. 12N
    nemotron-nano-12b-v2-vl:free
    ReasoningToolsVision14.9 intel · 28 t/s · 1.7s ttft
    14.9
    Intelligence
  13. 13N
    nemotron-nano-9b-v2:free
    ReasoningToolsJSON14.8 intel · 49 t/s · 922ms ttft
    14.8
    Intelligence
  14. 14M
    llama-3.3-70b-instruct:free
    Tools14.5 intel · 98 t/s · 205ms ttft
    14.5
    Intelligence
  15. 15L
    lfm-2.5-1.2b-thinking:free
    Reasoning8.1 intel · 126 t/s · 364ms ttft
    8.1
    Intelligence
  16. 16L
    lfm-2.5-1.2b-instruct:free
    8.0 intel · 95 t/s · 342ms ttft
    8.0
    Intelligence
  17. 17N
    nex-n2-pro:free
    ReasoningToolsJSON+125 t/s · 2.8s ttft · 262K ctx
    Intelligence
  18. 18N
    nemotron-3.5-content-safety:free
    ReasoningVision77 t/s · 257ms ttft · 128K ctx
    Intelligence
  19. 19P
    laguna-xs.2:free
    ReasoningTools88 t/s · 798ms ttft · 262K ctx
    Intelligence
  20. 20P
    laguna-m.1:free
    ReasoningTools12 t/s · 3.5s ttft · 262K ctx
    Intelligence
  21. 21G
    lyria-3-pro-preview
    JSONVision15 t/s · 11.1s ttft · 1.0M ctx
    Intelligence
  22. 22G
    lyria-3-clip-preview
    JSONVision8 t/s · 3.4s ttft · 1.0M ctx
    Intelligence
  23. 23C
    dolphin-mistral-24b-venice-edition:free
    JSON82 t/s · 440ms ttft · 33K ctx
    Intelligence
  24. 24M
    llama-3.2-3b-instruct:free
    72 t/s · 262ms ttft · 131K ctx
    Intelligence

Frequently asked

Is there a free LLM?

The most capable free LLM is Nemotron 3 Ultra (free), scoring 47.7 on the Intelligence Index at no per-token cost via OpenRouter. Gemma 4 31B (free) (39.2) and Nemotron 3 Super (free) (36.0) round out the top three.

What's a good alternative to Nemotron 3 Ultra (free)?

Gemma 4 31B (free) (39.2) is the closest alternative on this metric, followed by Nemotron 3 Super (free) (36.0). See the full ranking above for the tradeoffs.

By maker

All rankings