Weekly archive

New local AI models, Jul 15–21, 2024

13 open models released Jul 15–21, 2024, with how each runs on five reference devices.

  1. Llama 3.1 405B Instruct

    Meta · LLM · 406B ·

    50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  2. Llama 3.1 70B Instruct

    Meta · LLM · 71B ·

    66Good quality (B) · 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  3. ShieldGemma 27B

    Google · Decision · 27B ·

    66Good quality (B) · 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  4. ShieldGemma 9B

    Google · Decision · 9.2B ·

    64Good quality (B) · 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  5. ShieldGemma 2B

    Google · Decision · 2.6B ·

    58Good quality (B) · 58 of 100: how capable the model is at F16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  6. Mistral Nemo 12B Instruct

    Mistral AI · LLM · 12B ·

    51Fair quality (C) · 51 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  7. Llama 3.1 8B Instruct

    Meta · LLM · 8.0B ·

    48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  8. Codestral Mamba 7B

    Mistral AI · LLM · 7.3B ·

    36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  9. Mathstral 7B

    Mistral AI · LLM · 7.3B ·

    34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  10. Gemma 2 2B

    Google · LLM · 2.6B ·

    28Basic quality (D) · 28 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  11. SmolLM 1.7B Instruct

    Hugging Face · LLM · 1.7B ·

    14Basic quality (D) · 14 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  12. SmolLM 360M Instruct

    Hugging Face · LLM · 0.4B ·

    5Basic quality (D) · 5 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

  13. SmolLM 135M Instruct

    Hugging Face · LLM · 0.1B ·

    2Basic quality (D) · 2 of 100: how capable the model is at F16, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking.

    Runs on

    Memory, every quant and how to run it Check your own device

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.

Was this page helpful?