Request a Pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Phi-4 Mini Instruct

Microsoft

Strongest in Factual grounding (#103 of 104), weakest in Agents (#290 of 298). Above par in 3 of 9 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Opus 4.7 and ahead of GLM-4.6 and Qwen3 8B.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Safety47.1#246/329
Factual grounding26#103/104
Secure code54.5#117/270
Toxicity avoidance53.7#132/268
Harm refusal49.3#161/292
Jailbreak resistance51.4#164/268
Fairness44.5#198/292
Agents40#290/298
Knowledge work30#205/208

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Phi-4 Mini Instruct, left for the other.

OLMoE 1B-7B Instructwon 6 · lost 0
SmolLM 1.7B Instructwon 6 · lost 0
H2o Danube3 500M Chatwon 6 · lost 0
SmolLM2 1.7B Instructwon 6 · lost 0
RakutenAI 7B Instructwon 6 · lost 0
Qwen2.5 0.5B Instructwon 6 · lost 0
Mistral Instruct v0.1 (7B)won 5 · lost 1
OLMo 7B Instructwon 5 · lost 1
RakutenAI 7B Chatwon 5 · lost 1
DBRX Instructwon 5 · lost 1
PowerMoE 3Bwon 5 · lost 1
Phi 2won 5 · lost 1
Gemma 7B ITwon 5 · lost 1
PowerLM 3Bwon 5 · lost 1
Falcon Mamba 7B Instructwon 5 · lost 1
GLM-4 9B Chatwon 5 · lost 1
Nous Hermes 2 Mixtral 8x7B DPOwon 5 · lost 1
SmolLM2 360M Instructwon 5 · lost 1
CodeLlama 7B Instructwon 5 · lost 1
Qwen2 7B Instructwon 5 · lost 1
Magistral Smallwon 5 · lost 1
Aya 23 8Bwon 5 · lost 1
Laguna S 2.1won 5 · lost 1
Falcon H1 0.5B Instructwon 5 · lost 1
GPT-3.5 Turbowon 5 · lost 1
Qwen1.5 14B Chatwon 5 · lost 1
H2o Danube3 4B Chatwon 5 · lost 1
Zephyr 7B Betawon 5 · lost 1
Magistral Mediumwon 5 · lost 1
Jamba 1.6 Miniwon 5 · lost 1
LFM2 700Mwon 5 · lost 1
SmolLM 360M Instructwon 5 · lost 1
DeepSeek LLM 67B Chatwon 5 · lost 1
Mimo V2.6 Pro Ultraspeedwon 5 · lost 1
Qwen2.5 3B Instructwon 5 · lost 1
Grok 4.1 Fastwon 6 · lost 2
Jamba Mini 1.7won 5 · lost 2
Grok 3won 5 · lost 2
Command R+won 5 · lost 2
Jamba Large 1.7won 5 · lost 2
Command R7Bwon 5 · lost 2
Qwen3 30B A3B Instructwon 5 · lost 2
Grok 4won 5 · lost 2
LFM2 350Mwon 4 · lost 2
OLMo 2 7B Instructwon 4 · lost 2
Gemma 2B ITwon 4 · lost 2
SEA LION V1 7B ITwon 4 · lost 2
DeepSeek R1 Distill Qwen 7Bwon 4 · lost 2
DeepSeek R1 Distill Llama 8Bwon 4 · lost 2
Llama 3.1 70B Instructwon 4 · lost 2
NexusRaven V2 13Bwon 4 · lost 2
Jamba 1.5 Largewon 4 · lost 2
Mistral 7B Instruct v0.2won 4 · lost 2
Jamba 1.5 Miniwon 4 · lost 2
Granite 3.0 1B A400M Instructwon 4 · lost 2
GLM-4.5won 4 · lost 2
Jamba Instructwon 4 · lost 2
Sarvam 1won 4 · lost 2
SeaLLMs V3 7B Chatwon 4 · lost 2
Qwen 2.5 Instruct (7B)won 4 · lost 2
Magistral Small 1.2won 5 · lost 3
Grok 4.3won 5 · lost 3
NVIDIA Nemotron 3 Ultra 550B A55Bwon 5 · lost 3
Ling 3.0 Flash VLwon 5 · lost 3
Command Awon 5 · lost 3
Gemini 3.6 Flashwon 5 · lost 3
DeepSeek V4 Flashwon 5 · lost 3
MiMo V2.6 Flashwon 5 · lost 3
DeepSeek V3.1won 4 · lost 3
Granite 3.2 8B Instructwon 4 · lost 3
Gemma 4 31Bwon 5 · lost 4
DeepSeek V3won 5 · lost 4
Qwen3 8Bwon 5 · lost 4
GLM-4.6won 5 · lost 4
Mistral Instruct v0.3 (7B)won 3 · lost 3
Gemma 2 9B ITwon 3 · lost 3
Open Codestral Mambawon 3 · lost 3
Aya 23 35Bwon 3 · lost 3
LFM2 1.2Bwon 3 · lost 3
QwQ 32Bwon 3 · lost 3
Jamba 1.6 Largewon 3 · lost 3
DeepSeek Reasonerwon 3 · lost 3
WizardLM 2.8x22Bwon 3 · lost 3
Granite 3.1 2B Instructwon 3 · lost 3
Hunyuan A13B Instructwon 3 · lost 3
Granite 3.2 2B Instructwon 3 · lost 3
Granite 3.0 3B A800M Instructwon 3 · lost 3
Starling LM 7B Betawon 3 · lost 3
Granite 3.0 2B Instructwon 3 · lost 3
GLM-5 Turbowon 3 · lost 3
GPT-5.6 Astrawon 3 · lost 3
Krutrim 2 Instructwon 3 · lost 3
Ling 3.0 Flashwon 4 · lost 4
Grok Build 0.1won 4 · lost 4
DeepSeek R1won 4 · lost 5
Gemini 2.5 Prowon 4 · lost 5
Gemma 4 E2Bwon 3 · lost 4
Llama 3.2 3B Instructwon 3 · lost 4
GLM-5.2won 3 · lost 4
Llama 3.1 8B Instructwon 3 · lost 5
Gemma 4 E4Bwon 3 · lost 5
Gemma 3 4B ITwon 3 · lost 5
Gemma 4 12Bwon 3 · lost 5
Gemini 3.5 Flashwon 3 · lost 5
Qwen3.7 Maxwon 3 · lost 5
GLM-5.3 Flashwon 3 · lost 5
Gemini 2.5 Flashwon 3 · lost 5
Mixtral 8x22B Instructwon 2 · lost 4
Yi 1.5 9B Chatwon 2 · lost 4
Komodo 7B Basewon 2 · lost 4
Gemma 2 2B ITwon 2 · lost 4
IBM Granite 4.0 Microwon 2 · lost 4
Dolphin 2.5 Mixtral 8x7bwon 2 · lost 4
Gemini 1.5 Prowon 2 · lost 4
Gemma2 9B Cpt Sea Lionv3 Instructwon 2 · lost 4
Llama 3.1 405B Instructwon 2 · lost 4
DeepSeek Chatwon 2 · lost 4
LongWriter Glm4 9Bwon 2 · lost 4
Granite 3.1 3B A800M Instructwon 2 · lost 4
Phi-3 Medium 4K Instructwon 2 · lost 4
Sarvam 30Bwon 2 · lost 4
Qwen2.5 14B Instructwon 2 · lost 4
Qwen3 VL 235B A22B Instructwon 2 · lost 4
Gemini 2.0 Flashwon 2 · lost 4
Phi 3 Medium 128k Instructwon 2 · lost 4
Granite 3.0 8B Instructwon 2 · lost 4
Llama 3 8B Instructwon 2 · lost 4
EAI Llama 3.1 8B Instruct SageAlignwon 2 · lost 4
Phi 3.5 Mini Instructwon 2 · lost 4
K2 Chatwon 2 · lost 4
Llama 3.1 Tulu 3 8B SFTwon 2 · lost 4
Phi-3 Mini 4K Instructwon 2 · lost 4
Llama 3.1 Tulu 3 8B DPOwon 2 · lost 4
Llama 3.1 Tulu 3 8Bwon 2 · lost 4
EAI SageAlign 8B LoRAwon 2 · lost 4
Llama 3.2 Vision (11B)won 2 · lost 4
Llama 2 13B Chatwon 2 · lost 4
Gemma 4 26B A4Bwon 3 · lost 6
DeepSeek V3.2won 3 · lost 6
Kimi K2.6won 3 · lost 6
Mistral Small 3 24B Instructwon 2 · lost 5
Amazon Nova Litewon 2 · lost 5
Llama 3.2 1B Instructwon 2 · lost 5
Granite 3.1 8B Instructwon 2 · lost 5
Gemma 3 12B ITwon 2 · lost 6
Amazon Nova Microwon 2 · lost 6
GPT OSS 20Bwon 2 · lost 6
Amazon Nova Prowon 2 · lost 6
NVIDIA Nemotron 3 Super 120B A12Bwon 2 · lost 6
GPT-6 Lunawon 2 · lost 6
Mistral Large 4won 2 · lost 6
Gemini 3.8 Flashwon 2 · lost 6
Kimi K2.7 Codewon 2 · lost 6
MiMo V2.5 Prowon 2 · lost 6
Kimi K2.5won 2 · lost 6
Qwen3 235B A22B Instructwon 2 · lost 6
Kimi K2won 2 · lost 6
MiniMax M3won 2 · lost 6
Llama 4 Scout Instructwon 2 · lost 7
Llama 3.3 70B Instructwon 2 · lost 7
Mercury 2won 2 · lost 7
GPT OSS 120Bwon 2 · lost 7
DeepSeek V4 Prowon 2 · lost 7
GPT-6 Solwon 2 · lost 7
Reflection Llama 3.1 70Bwon 1 · lost 5
Qwen2.5 32B Instructwon 1 · lost 5
Gemini 1.5 Flashwon 1 · lost 5
Qwen2 VL 72Bwon 1 · lost 5
Claude 3 Haikuwon 1 · lost 5
Qwen2 57B A14B Instructwon 1 · lost 5
Smaug Llama 3 70B Instructwon 1 · lost 5
GPT OSS Safeguard 20Bwon 1 · lost 5
SeaLLM 7B V2won 1 · lost 5
Apertus 70B Instructwon 1 · lost 5
Granite 4.0 H Tinywon 1 · lost 5
Phi-3 Small 8K Instructwon 1 · lost 5
Llama 3 70B Instructwon 1 · lost 5
Qwen3.5 27Bwon 1 · lost 5
Llama 3.1 Nemotron 70B Instructwon 1 · lost 5
Titan Text Express V1won 1 · lost 5
Llama 2 70B Chatwon 1 · lost 5
Llama 3 8B Instruct RRwon 1 · lost 5
Nova Premierwon 1 · lost 5
Fugu Ultra V2won 1 · lost 5
Sarvam 105Bwon 1 · lost 5
Llama 3 8B Instruct MopeyMulewon 1 · lost 5
Llama 3.2 Vision (90B)won 1 · lost 5
Qwen2.5 72B Instructwon 1 · lost 5
Claude 3.5 Sonnetwon 1 · lost 5
Claude Opus 4won 1 · lost 6
Claude 3.7 Sonnetwon 1 · lost 6
O3won 1 · lost 6
Claude Sonnet 4.6won 1 · lost 6
O3 Miniwon 1 · lost 7
Gemma 3 27B ITwon 1 · lost 7
GPT-4.1 nanowon 1 · lost 7
GPT-4won 1 · lost 7
Claude 3.5 Haikuwon 1 · lost 7
Qwen3.5 122B A10Bwon 1 · lost 7
Qwen3.5 397B A17Bwon 1 · lost 7
GPT-4.1 miniwon 1 · lost 7
Grok 4.5won 1 · lost 7
GLM-5.1won 1 · lost 7
Qwen3.8 Maxwon 1 · lost 7
GPT-5.6 Terrawon 1 · lost 7
O4 Miniwon 1 · lost 7
GPT-4.1won 1 · lost 7
Muse Spark 1.1won 1 · lost 7
Claude Opus 4.8won 1 · lost 7
GPT-5.6 Lunawon 1 · lost 7
MiMo V2.6 Prowon 1 · lost 7
Llama 4 Maverick Instructwon 1 · lost 8
Qwen3.5 35B A3Bwon 1 · lost 8
GPT-5 miniwon 1 · lost 8
GPT-5.1won 1 · lost 8
Claude Haiku 4.5won 1 · lost 8
GPT-5.4won 1 · lost 8
GPT-5.5won 1 · lost 8
Claude Sonnet 4.5won 1 · lost 8
GPT-5.6 Solwon 1 · lost 8
Granite 4.0 H Microwon 0 · lost 6
Qwen Maxwon 0 · lost 6
Phi 3.5 MoE Instructwon 0 · lost 6
Granite 4.0 H Smallwon 0 · lost 6
Qwen2 72B Instructwon 0 · lost 6
Claude 3 Sonnetwon 0 · lost 6
Cogito 671B V2 P1won 0 · lost 6
Foundation Sec 8B Instructwon 0 · lost 6
Gemma 2 27B ITwon 0 · lost 6
O1 Miniwon 0 · lost 6
Meta SecAlign 8Bwon 0 · lost 6
Apriel 1.5 15B Thinkerwon 0 · lost 6
GPT-4 Turbowon 0 · lost 6
Mistral NeMo Minitron 8B Instructwon 0 · lost 6
O1won 0 · lost 6
MiniMax M2won 0 · lost 6
Claude 3 Opuswon 0 · lost 6
Qwen3 Next 80B A3B Thinkingwon 0 · lost 6
Aya Expanse 32Bwon 0 · lost 7
Aya Expanse 8Bwon 0 · lost 7
Inklingwon 0 · lost 8
GPT-4o miniwon 0 · lost 8
Phi 4won 0 · lost 8
Hy3won 0 · lost 8
GPT-5 nanowon 0 · lost 8
Grok 4.6won 0 · lost 8
GPT-5.2won 0 · lost 8
Claude Sonnet 5won 0 · lost 8
Grok 4.7won 0 · lost 8
Claude Opus 5won 0 · lost 8
GPT-4owon 0 · lost 9
Claude Sonnet 4won 0 · lost 9
MiniMax M2.7won 0 · lost 9
GPT-5won 0 · lost 9
Claude Opus 4.7won 0 · lost 9
Claude Opus 5.5won 0 · lost 9

§ 3 · Sources

Where the numbers come from

3 publications, 8 figures. Every one links to the page it was read from.

GDPval-AA ↗1 measureread 2026-11-10
GDPval-AA -401
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-11-10
Enkrypt · Jailbreak risk 11.6%Enkrypt · Harmful content risk 39.4%Enkrypt · CBRN risk 13.7%Enkrypt · Toxicity risk 3.6%Enkrypt · Bias risk 86.3%Enkrypt · Insecure code risk 24%
Vectara Hallucination Leaderboard ↗1 measureread 2026-11-10
Vectara · Factual consistency 76.5%

Badge

PublicAI Index badge for Phi-4 Mini Instruct[![PublicAI Index](https://publicai.io/model-index/badge?model=phi-4-mini-instruct)](https://publicai.io/model-index/m/phi-4-mini-instruct)