model family

Phi 4

Published by microsoft. 13 members have been measured on real machines. Every number below was signed by the machine that produced it — none of it is from a model card.

what it takes, what it returns
text → text

Declared by the network’s catalog, not measured — a signature says what this family would be asked. Whether it answered is the signed capability flags on each member.

13 of 60 builds measured · 47 still open
what each build proved
mini · Q3_K_L · bartowski
mini · IQ2_M
mini · Q3_K_S · MaziyarPanahi
mini-reasoning · IQ1_S
mini-reasoning · Q2_K_L
mini-reasoning · Q3_K_M
mini-reasoning · Q2_K
mini-reasoning · Q4_K_M
mini · Q3_K_S · bartowski
mini · Q4_K_M · bartowski
mini · Q3_K_L · MaziyarPanahi
mini · IQ2_XS
mini · Q4_K_M · unsloth
mini · Q3_K_L · bartowski · Structured output — provedmini · Q3_K_L · bartowski · Sustained output — provedmini · Q3_K_L · bartowski · Extended thinking — asked, failedmini · Q3_K_L · bartowski · Writes code — provedmini · Q3_K_L · bartowski · Reasons about code — asked, failedmini · Q3_K_L · bartowski · Synthesises sources — provedmini · Q3_K_L · bartowski · Follows instructions — provedmini · IQ2_M · Structured output — provedmini · IQ2_M · Sustained output — provedmini · IQ2_M · Extended thinking — asked, failedmini · IQ2_M · Writes code — provedmini · IQ2_M · Reasons about code — asked, failedmini · IQ2_M · Synthesises sources — asked, failedmini · IQ2_M · Follows instructions — provedmini · Q3_K_S · MaziyarPanahi · Structured output — provedmini · Q3_K_S · MaziyarPanahi · Sustained output — provedmini · Q3_K_S · MaziyarPanahi · Extended thinking — asked, failedmini · Q3_K_S · MaziyarPanahi · Writes code — provedmini · Q3_K_S · MaziyarPanahi · Reasons about code — asked, failedmini · Q3_K_S · MaziyarPanahi · Synthesises sources — provedmini · Q3_K_S · MaziyarPanahi · Follows instructions — asked, failedmini-reasoning · IQ1_S · Structured output — asked, failedmini-reasoning · IQ1_S · Sustained output — asked, failedmini-reasoning · IQ1_S · Extended thinking — asked, failedmini-reasoning · IQ1_S · Writes code — asked, failedmini-reasoning · IQ1_S · Reasons about code — asked, failedmini-reasoning · IQ1_S · Synthesises sources — asked, failedmini-reasoning · IQ1_S · Follows instructions — asked, failedmini-reasoning · Q2_K_L · Structured output — provedmini-reasoning · Q2_K_L · Sustained output — never askedmini-reasoning · Q2_K_L · Extended thinking — provedmini-reasoning · Q2_K_L · Writes code — asked, failedmini-reasoning · Q2_K_L · Reasons about code — provedmini-reasoning · Q2_K_L · Synthesises sources — never askedmini-reasoning · Q2_K_L · Follows instructions — asked, failedmini-reasoning · Q3_K_M · Structured output — provedmini-reasoning · Q3_K_M · Sustained output — provedmini-reasoning · Q3_K_M · Extended thinking — provedmini-reasoning · Q3_K_M · Writes code — asked, failedmini-reasoning · Q3_K_M · Reasons about code — provedmini-reasoning · Q3_K_M · Synthesises sources — provedmini-reasoning · Q3_K_M · Follows instructions — asked, failedmini-reasoning · Q2_K · Structured output — provedmini-reasoning · Q2_K · Sustained output — provedmini-reasoning · Q2_K · Extended thinking — provedmini-reasoning · Q2_K · Writes code — asked, failedmini-reasoning · Q2_K · Reasons about code — provedmini-reasoning · Q2_K · Synthesises sources — provedmini-reasoning · Q2_K · Follows instructions — asked, failedmini-reasoning · Q4_K_M · Structured output — provedmini-reasoning · Q4_K_M · Sustained output — asked, failedmini-reasoning · Q4_K_M · Extended thinking — provedmini-reasoning · Q4_K_M · Writes code — asked, failedmini-reasoning · Q4_K_M · Reasons about code — provedmini-reasoning · Q4_K_M · Synthesises sources — provedmini-reasoning · Q4_K_M · Follows instructions — asked, failedmini · Q3_K_S · bartowski · Structured output — never askedmini · Q3_K_S · bartowski · Sustained output — never askedmini · Q3_K_S · bartowski · Extended thinking — never askedmini · Q3_K_S · bartowski · Writes code — never askedmini · Q3_K_S · bartowski · Reasons about code — never askedmini · Q3_K_S · bartowski · Synthesises sources — never askedmini · Q3_K_S · bartowski · Follows instructions — never askedmini · Q4_K_M · bartowski · Structured output — never askedmini · Q4_K_M · bartowski · Sustained output — never askedmini · Q4_K_M · bartowski · Extended thinking — never askedmini · Q4_K_M · bartowski · Writes code — never askedmini · Q4_K_M · bartowski · Reasons about code — never askedmini · Q4_K_M · bartowski · Synthesises sources — never askedmini · Q4_K_M · bartowski · Follows instructions — never askedmini · Q3_K_L · MaziyarPanahi · Structured output — never askedmini · Q3_K_L · MaziyarPanahi · Sustained output — never askedmini · Q3_K_L · MaziyarPanahi · Extended thinking — never askedmini · Q3_K_L · MaziyarPanahi · Writes code — never askedmini · Q3_K_L · MaziyarPanahi · Reasons about code — never askedmini · Q3_K_L · MaziyarPanahi · Synthesises sources — never askedmini · Q3_K_L · MaziyarPanahi · Follows instructions — never askedmini · IQ2_XS · Structured output — provedmini · IQ2_XS · Sustained output — provedmini · IQ2_XS · Extended thinking — asked, failedmini · IQ2_XS · Writes code — asked, failedmini · IQ2_XS · Reasons about code — provedmini · IQ2_XS · Synthesises sources — asked, failedmini · IQ2_XS · Follows instructions — asked, failedmini · Q4_K_M · unsloth · Structured output — never askedmini · Q4_K_M · unsloth · Sustained output — never askedmini · Q4_K_M · unsloth · Extended thinking — never askedmini · Q4_K_M · unsloth · Writes code — never askedmini · Q4_K_M · unsloth · Reasons about code — never askedmini · Q4_K_M · unsloth · Synthesises sources — never askedmini · Q4_K_M · unsloth · Follows instructions — never asked
Structured output
Sustained output
Extended thinking
Writes code
Reasons about code
Synthesises sources
Follows instructions
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q3_K_Lmini · Q3_K_L · 1 machine
8,192recalled
—served
32.1best tok/s
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ2_Mmini · IQ2_M · 2 machines
4,096recalled
—served
37.3best tok/s
37.3 tok/s · unnamed machine1.1 tok/s · unnamed machine
1.1 → 37.3 tok/s across 2 machines · ×35.3 spreadunnamed machine
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_Smini · Q3_K_S · 2 machines
4,096recalled
—served
27.9best tok/s
27.9 tok/s · unnamed machine2.2 tok/s · unnamed machine
2.2 → 27.9 tok/s across 2 machines · ×12.9 spreadunnamed machine
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ1_Smini-reasoning · IQ1_S · 5 machines · 3.84B
—recalled
—served
188.9best tok/s
188.9 tok/s · NVIDIA · Bhubaneswar26.4 tok/s · unnamed machine16.1 tok/s · unnamed machine5.5 tok/s · unnamed machine4.0 tok/s · x86 · CPU · Hyderabad
4.0 → 188.9 tok/s across 5 machines · ×46.8 spreadNVIDIAunnamed machinex86 · CPU
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:Q2_K_Lmini-reasoning · Q2_K_L · 3 machines
—recalled
—served
6.8best tok/s
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:Q3_K_Mmini-reasoning · Q3_K_M · 3 machines
—recalled
—served
31.9best tok/s
31.9 tok/s · unnamed machine20.2 tok/s · unnamed machine0.0 tok/s · unnamed machine
0.0 → 31.9 tok/s across 3 machines · ×1447.5 spreadunnamed machine
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:Q2_Kmini-reasoning · Q2_K · 2 machines
—recalled
—served
37.7best tok/s
37.7 tok/s · unnamed machine9.8 tok/s · unnamed machine
9.8 → 37.7 tok/s across 2 machines · ×3.8 spreadunnamed machine
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:Q4_K_Mmini-reasoning · Q4_K_M · 2 machines
—recalled
—served
6.5best tok/s
6.5 tok/s · unnamed machine1.9 tok/s · unnamed machine
1.9 → 6.5 tok/s across 2 machines · ×3.5 spreadunnamed machine
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q3_K_Smini · Q3_K_S · 1 machine
—recalled
—served
28.2best tok/s
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q4_K_Mmini · Q4_K_M · 1 machine
—recalled
—served
40.6best tok/s
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_Lmini · Q3_K_L · 1 machine
—recalled
—served
2.2best tok/s
hf.co/Mungert/Phi-4-mini-instruct.gguf:IQ2_XSmini · IQ2_XS · 1 machine
—recalled
—served
8.5best tok/s
hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q4_K_Mmini · Q4_K_M · 1 machine
—recalled
—served
2.2best tok/s

Builds this network knows of

Declared, not measured — this is what the catalog says exists and what it would cost a machine to hold. Nothing here is signed, and none of it claims the build works.

how much memory does your machine have

49 of 60 builds fit a 16 GB machine — budgeting 70% of it, or 11.2 GiB, because the cache and the runtime want the rest.

  • optimum-intel-internal-testing/tiny-random-phi-4-multimodalFP32vllmderived by optimum-intel-internal-testing
    0.0 GiB
  • hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ1_SIQ1_Sollamameasured herequantized by unsloth
    1.1 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ2_MIQ2_Mollamameasured herederived by bartowski
    1.4 GiB
  • hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ3_XXSIQ3_XXSollamaquantized by unsloth
    1.6 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ3_XXSIQ3_XXSollamaderived by bartowski
    1.6 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q2_KQ2_Kollamaquantized by MaziyarPanahi
    1.6 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_SQ3_K_Sollamameasured herequantized by MaziyarPanahi
    1.8 GiB
  • mlx-community/Phi-4-mini-instruct-4bit4BITmlxquantized by mlx-community
    2.0 GiB
  • lmstudio-community/Phi-4-mini-reasoning-MLX-4bit4BITmlxquantized by lmstudio-community
    2.0 GiB
  • tbhrc/unsloth_phi_4_mini_instruct_4bit4BITmlxderived by tbhrc
    2.0 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ4_XSIQ4_XSollamaderived by bartowski
    2.1 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_LQ3_K_Lollamameasured herequantized by MaziyarPanahi
    2.1 GiB
  • hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q3_K_LQ3_K_Lollamaquantized by lmstudio-community
    2.1 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q4_0Q4_0ollamaderived by bartowski
    2.2 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q4_K_SQ4_K_Sollamaquantized by MaziyarPanahi
    2.2 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q4_K_MQ4_K_Mollamaquantized by MaziyarPanahi
    2.3 GiB
  • hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q4_K_MQ4_K_Mollamaquantized by lmstudio-community
    2.3 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q4_K_MQ4_K_Mollamameasured herederived by bartowski
    2.3 GiB
  • hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q6_KQ6_Kollamaquantized by lmstudio-community
    2.9 GiB
  • unsloth/Phi-4-mini-instruct-unsloth-bnb-4bitBITSANDBYTESvllmderived by unsloth
    3.0 GiB
  • mlx-community/Phi-4-mini-instruct-8bit8BITmlxquantized by mlx-community
    3.8 GiB
  • hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q8_0Q8_0ollamaquantized by MaziyarPanahi
    3.8 GiB
  • hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q8_0Q8_0ollamaquantized by lmstudio-community
    3.8 GiB
  • hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q8_0Q8_0ollamaderived by bartowski
    3.8 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:IQ2_SIQ2_Sollamaderived by bartowski
    4.4 GiB
  • hf.co/microsoft/phi-4-gguf:Q2_KQ2_Kollamapublished by microsoft
    5.2 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:IQ3_XXSIQ3_XXSollamaderived by bartowski
    5.4 GiB
  • hf.co/microsoft/phi-4-gguf:IQ3_XXSIQ3_XXSollamapublished by microsoft
    5.7 GiB
  • hf.co/microsoft/phi-4-gguf:IQ3_SIQ3_Sollamapublished by microsoft
    6.1 GiB
  • nvidia/Phi-4-multimodal-instruct-NVFP4NVFP4vllmquantized by nvidia
    6.8 GiB
  • microsoft/Phi-4-mini-instructBF16vllmpublished by microsoft
    7.1 GiB
  • microsoft/Phi-4-mini-reasoningBF16vllmpublished by microsoft
    7.1 GiB
  • mlx-community/Phi-4-mini-instruct-mlx-fp16BF16mlxderived by mlx-community
    7.1 GiB
  • unsloth/Phi-4-mini-instructBF16vllmderived by unsloth
    7.1 GiB
  • hf.co/unsloth/Phi-4-mini-instruct-GGUF:BF16BF16ollamaquantized by unsloth
    7.2 GiB
  • hf.co/unsloth/Phi-4-mini-reasoning-GGUF:BF16BF16ollamaquantized by unsloth
    7.2 GiB
  • hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q3_K_LQ3_K_Lollamaquantized by lmstudio-community
    7.4 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q3_K_LQ3_K_Lollamaderived by bartowski
    7.4 GiB
  • mlx-community/phi-4-4bit4BITmlxquantized by mlx-community
    7.7 GiB
  • mlx-community/Unsloth-Phi-4-4bit4BITmlxderived by mlx-community
    7.7 GiB
  • lmstudio-community/Phi-4-reasoning-MLX-4bit4BITmlxquantized by lmstudio-community
    7.7 GiB
  • lmstudio-community/Phi-4-reasoning-plus-MLX-4bit4BITmlxquantized by lmstudio-community
    7.7 GiB
  • hf.co/microsoft/phi-4-gguf:Q4_0Q4_0ollamapublished by microsoft
    7.8 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q4_0Q4_0ollamaderived by bartowski
    7.8 GiB
  • hf.co/MaziyarPanahi/phi-4-GGUF:Q4_K_MQ4_K_Mollamaquantized by MaziyarPanahi
    8.4 GiB
  • hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q4_K_MQ4_K_Mollamaquantized by lmstudio-community
    8.4 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q4_K_MQ4_K_Mollamaderived by bartowski
    8.4 GiB
  • unsloth/phi-4-unsloth-bnb-4bitBITSANDBYTESvllmderived by unsloth
    9.7 GiB
  • microsoft/Phi-4-multimodal-instructBF16vllmpublished by microsoft
    10.4 GiB
  • hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q6_KQ6_Kollamaquantized by lmstudio-community
    11.2 GiB
  • mlx-community/phi-4-8bit8BITmlxquantized by mlx-community
    14.5 GiB
  • hf.co/microsoft/phi-4-gguf:Q8_0Q8_0ollamapublished by microsoft
    14.5 GiB
  • hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q8_0Q8_0ollamaquantized by lmstudio-community
    14.5 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q8_0Q8_0ollamaderived by bartowski
    14.5 GiB
  • RedHatAI/phi-4-FP8-dynamicFP8vllmquantized by RedHatAI
    14.6 GiB
  • microsoft/Phi-4-reasoningBF16vllmpublished by microsoft
    27.3 GiB
  • microsoft/Phi-4-reasoning-plusBF16vllmpublished by microsoft
    27.3 GiB
  • hf.co/microsoft/phi-4-gguf:BF16BF16ollamapublished by microsoft
    27.3 GiB
  • hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:BF16BF16ollamaderived by bartowski
    27.3 GiB
  • microsoft/Phi-4-reasoning-vision-15BBF16vllmpublished by microsoft
    28.2 GiB

Sizes are what the catalog declares, not what a machine measured — fitting on disk is necessary and not sufficient. Nobody here has run these on your machine, and this narrows the candidates rather than certifying one.

the same builds, by member
Phi 4 15B
microsoft/Phi-4-reasoning-vision-15BBF16vllm28.2 GiB on diskpublished by microsoft
Phi 4 BASE
hf.co/microsoft/phi-4-gguf:Q2_KQ2_Kollama5.2 GiB on diskpublished by microsoft
hf.co/microsoft/phi-4-gguf:IQ3_XXSIQ3_XXSollama5.7 GiB on diskpublished by microsoft
hf.co/microsoft/phi-4-gguf:IQ3_SIQ3_Sollama6.1 GiB on diskpublished by microsoft
mlx-community/phi-4-4bit4BITmlx7.7 GiB on diskquantized by mlx-community
hf.co/microsoft/phi-4-gguf:Q4_0Q4_0ollama7.8 GiB on diskpublished by microsoft
hf.co/MaziyarPanahi/phi-4-GGUF:Q4_K_MQ4_K_Mollama8.4 GiB on diskquantized by MaziyarPanahi
mlx-community/phi-4-8bit8BITmlx14.5 GiB on diskquantized by mlx-community
hf.co/microsoft/phi-4-gguf:Q8_0Q8_0ollama14.5 GiB on diskpublished by microsoft
RedHatAI/phi-4-FP8-dynamicFP8vllm14.6 GiB on diskquantized by RedHatAI
hf.co/microsoft/phi-4-gguf:BF16BF16ollama27.3 GiB on diskpublished by microsoft
Phi 4 BASE · mlx-community
mlx-community/Unsloth-Phi-4-4bit4BITmlx7.7 GiB on diskderived by mlx-community
Phi 4 BASE · unsloth
unsloth/phi-4-unsloth-bnb-4bitBITSANDBYTESvllm9.7 GiB on diskderived by unsloth
Phi 4 MINI
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q2_KQ2_Kollama1.6 GiB on diskquantized by MaziyarPanahi
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_SQ3_K_Sollama1.8 GiB on diskquantized by MaziyarPanahi
mlx-community/Phi-4-mini-instruct-4bit4BITmlx2.0 GiB on diskquantized by mlx-community
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q3_K_LQ3_K_Lollama2.1 GiB on diskquantized by MaziyarPanahi
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q4_K_SQ4_K_Sollama2.2 GiB on diskquantized by MaziyarPanahi
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q4_K_MQ4_K_Mollama2.3 GiB on diskquantized by MaziyarPanahi
mlx-community/Phi-4-mini-instruct-8bit8BITmlx3.8 GiB on diskquantized by mlx-community
hf.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF:Q8_0Q8_0ollama3.8 GiB on diskquantized by MaziyarPanahi
microsoft/Phi-4-mini-instructBF16vllm7.1 GiB on diskpublished by microsoft
hf.co/unsloth/Phi-4-mini-instruct-GGUF:BF16BF16ollama7.2 GiB on diskquantized by unsloth
Phi 4 MINI-REASONING
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ1_SIQ1_Sollama1.1 GiB on diskquantized by unsloth
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ3_XXSIQ3_XXSollama1.6 GiB on diskquantized by unsloth
lmstudio-community/Phi-4-mini-reasoning-MLX-4bit4BITmlx2.0 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q3_K_LQ3_K_Lollama2.1 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q4_K_MQ4_K_Mollama2.3 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q6_KQ6_Kollama2.9 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-mini-reasoning-GGUF:Q8_0Q8_0ollama3.8 GiB on diskquantized by lmstudio-community
microsoft/Phi-4-mini-reasoningBF16vllm7.1 GiB on diskpublished by microsoft
hf.co/unsloth/Phi-4-mini-reasoning-GGUF:BF16BF16ollama7.2 GiB on diskquantized by unsloth
Phi 4 MINI · bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ2_MIQ2_Mollama1.4 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ3_XXSIQ3_XXSollama1.6 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:IQ4_XSIQ4_XSollama2.1 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q4_0Q4_0ollama2.2 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q4_K_MQ4_K_Mollama2.3 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF:Q8_0Q8_0ollama3.8 GiB on diskderived by bartowski
Phi 4 MINI · mlx-community
mlx-community/Phi-4-mini-instruct-mlx-fp16BF16mlx7.1 GiB on diskderived by mlx-community
Phi 4 MINI · tbhrc
tbhrc/unsloth_phi_4_mini_instruct_4bit4BITmlx2.0 GiB on diskderived by tbhrc
Phi 4 MINI · unsloth
unsloth/Phi-4-mini-instruct-unsloth-bnb-4bitBITSANDBYTESvllm3.0 GiB on diskderived by unsloth
unsloth/Phi-4-mini-instructBF16vllm7.1 GiB on diskderived by unsloth
Phi 4 MULTIMODAL
nvidia/Phi-4-multimodal-instruct-NVFP4NVFP4vllm6.8 GiB on diskquantized by nvidia
microsoft/Phi-4-multimodal-instructBF16vllm10.4 GiB on diskpublished by microsoft
Phi 4 REASONING
lmstudio-community/Phi-4-reasoning-MLX-4bit4BITmlx7.7 GiB on diskquantized by lmstudio-community
microsoft/Phi-4-reasoningBF16vllm27.3 GiB on diskpublished by microsoft
Phi 4 REASONING-PLUS
hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q3_K_LQ3_K_Lollama7.4 GiB on diskquantized by lmstudio-community
lmstudio-community/Phi-4-reasoning-plus-MLX-4bit4BITmlx7.7 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q4_K_MQ4_K_Mollama8.4 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q6_KQ6_Kollama11.2 GiB on diskquantized by lmstudio-community
hf.co/lmstudio-community/Phi-4-reasoning-plus-GGUF:Q8_0Q8_0ollama14.5 GiB on diskquantized by lmstudio-community
microsoft/Phi-4-reasoning-plusBF16vllm27.3 GiB on diskpublished by microsoft
Phi 4 REASONING-PLUS · bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:IQ2_SIQ2_Sollama4.4 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:IQ3_XXSIQ3_XXSollama5.4 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q3_K_LQ3_K_Lollama7.4 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q4_0Q4_0ollama7.8 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q4_K_MQ4_K_Mollama8.4 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:Q8_0Q8_0ollama14.5 GiB on diskderived by bartowski
hf.co/bartowski/microsoft_Phi-4-reasoning-plus-GGUF:BF16BF16ollama27.3 GiB on diskderived by bartowski
Phi 4 TINY-RANDOM-MULTIMODAL · optimum-intel-internal-testing
optimum-intel-internal-testing/tiny-random-phi-4-multimodalFP32vllm24 MiB on diskderived by optimum-intel-internal-testing
ask your own gram

Your concierge can answer this from the network rather than from this page. The ask carries the operator id, so it resolves to a capability and an engine — not to a name somebody has to recognise.

What has this network measured about the Phi 4 family (epn.family.phi-4), and can my gram run any of it? See /inferences/phi-4

open a gramx with this ask →epn.family.phi-4