Phi 3 Mini 128K Instruct
Apr 2024
Phi 3 Mini 128K InstructCompare inon
Benchmarks of Phi 3 Mini 128K Instruct14
BenchmarkByScoreRank
No benchmark matches
—per task?
self-host onlyprice
—output speed?
131Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-5 Minihighwins 80%—$0.053
OpenGemma 3 12Bunstatedwins 81%—$0.05 / $0.10Memory (4-bit): 1.9 GB → 6.1 GBBetterGPT-6 Lunamaxwins 87%—$0.068Function calling only with none · Chat Completions+3BetterGPT-5.6 Lunaxhighwins 87%—$0.085BetterGPT-5.6 Terraxhighwins 88%—$0.63Prompts with >272K input tokensModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 80%—$0.053GPT-6 Lunamaxwins 87%—$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 87%—$0.085GPT-5.6 Terraxhighwins 88%—$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 89%—$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Gemma 3 12BunstatedOpenwins 81%—$0.05 / $0.10Memory (4-bit): 1.9 GB → 6.1 GBGemma 3n E4B InstructunstatedOpenwins 77%——Context: 131K → 32K+1Gemma 3 4BunstatedOpenwins 75%—$0.04 / $0.05Memory (4-bit): 1.9 GB → 2.2 GB
Granite 4.1 8BunstatedOpenwins 73%——Memory (4-bit): 1.9 GB → 4.4 GBGemma 2 9B ItOpenwins 73%—$0.03 / $0.06Context: 131K → 8K+1ModelScoreCostPer taskWhat changes
Phi 4unstatedOpenwins 67%—$0.056 / $0.11Context: 131K → 16K+1Phi 3 Medium 4K InstructOpenwins 59%—$0.14 / $0.14Context: 131K → 4K+1Phi 3 Small 8K InstructOpenwins 57%——Context: 131K → 8K+1Phi 3 Mini 4K InstructOpensame—$0.40 / $0.80Context: 131K → 4KPhi 4 Reasoning PlusOpenn/a—$0.07 / $0.35Context: 131K → 33K+1Phi 4 ReasoningOpenn/a——Context: 131K → 33K+1Phi 3.5 Mini InstructOpenn/a——Phi 3 Medium 128K InstructOpenn/a——Memory (4-bit): 1.9 GB → 7 GBAll Microsoft models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · full quality✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceMITfree for commercial use
Benchmarks
all categoriesChat & Knowledge
Arena Text · overall1,130#191
Arena Text · Chinese1,016.3#178
Arena Text · English1,082.9#186
Arena Text · French1,039#147
Arena Text · German1,006#152
Arena Text · instruction following1,022.7#186
Arena Text · Japanese899.1#137
Arena Text · Korean856#144
Arena Text · non-English1,000.4#184
Arena Text · Russian1,003.9#182
Arena Text · Spanish1,058.4#153
Arena Text · overall · full cohort1,050.3#185
Writing & Language
Arena Text · creative writing1,086#191
FAQ
How much does Phi 3 Mini 128K Instruct cost?›
There is no per-unit API price: self-host only.
Is Phi 3 Mini 128K Instruct free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Phi 3 Mini 128K Instruct, and what is it best at?›
#191 of 196 in Chat & Knowledge, #191 of 195 in Writing & Language.
Is Phi 3 Mini 128K Instruct better than Mistral 7B Instruct?›
On Arena Text · creative writing, Phi 3 Mini 128K Instruct scores 1,086 vs 1,091 — about the same.
What is the context window of Phi 3 Mini 128K Instruct?›
131K tokens.
When was Phi 3 Mini 128K Instruct released? Is there a newer version?›
Released 2024-04-22. It is the current version in its line.
Is Phi 3 Mini 128K Instruct open source?›
Yes, the weights are open (MIT licence).
Can I run Phi 3 Mini 128K Instruct locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Phi 3 Mini 128K Instruct read images?›
No, it accepts text only.
What is Phi 3 Mini 128K Instruct?›
Phi 3 Mini 128K Instruct is an open-weights language model from Microsoft, released Apr 2024.
What are the best alternatives to Phi 3 Mini 128K Instruct?›
GPT-5 Mini (better), Gemma 3 12B (open weights).