Skip to content
Model Pareto

Phi-4-reasoning-plus

Microsoft AIOpen weightshuggingface.co ↗
Compare with nearest rivalsCompare with…
Input
—
Output
—
Speed
~48tok/sest.
Context
33Ktokens

14B dense reasoning model (Phi-4 + SFT + RL). No API price found (not on OpenRouter).

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.

Hardware
1× B200 180GB
BF16 weights
Throughput
~7,374 tok/s
many requests batched
Price
$0.12–1.08 /M tok
busy → light use
On your own machine
RTX 5090 32GB (int4, ~170 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • BF16 weights (14.7B params) + 25% KV-cache headroom ≈ 37 GB
  • 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~7,374 output tok/s aggregate at batch 211 (bandwidth-bound)
  • Blended 3:1 input:output; prefill ~36,161 tok/s
  • Low = 75% utilization at the low GPU price; high = 20% at the high price

Benchmark scores

~ italic, dashed = no published score yet, estimated from related benchmarks.

Coding

SWE-bench VerifiedEstimated
~#5 of 14 (estimated)~78.8%
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. #4MiniMax-M380.5%
  4. ~#5Phi-4-reasoning-plusEstimated78.8%
  5. #5Hy378.0%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#28 of 31 (estimated)~45.3%
  1. #1DeepSeek V4.1 Flash74.2%
  2. #2Grok 4.772.6%
  3. #27Qwen3.8 Max (0902)51.0%
  4. ~#28Phi-4-reasoning-plusEstimated45.3%
  5. #28Qwen3.8 27B42.2%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#19 of 84 (estimated)~18.1%
  1. #1Claude Sonnet 5.563.6%
  2. #2Claude Opus 5.559.6%
  3. #18Gemini 3.8 Flash19.7%
  4. ~#19Phi-4-reasoning-plusEstimated18.1%
  5. #19Grok 4.617.2%
No published score — estimated from related benchmarks
SciCodeEstimated
~#39 of 96 (estimated)~46.3%
  1. #1Claude Opus 5.566.9%
  2. #2Claude Fable 5.163.1%
  3. #38Qwen3.8 27B46.6%
  4. ~#39Phi-4-reasoning-plusEstimated46.3%
  5. #39Qwen3.7 Plus46.1%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~#10 of 15 (estimated)~48.0
  1. #1Claude Opus 5.566.0
  2. #2Claude Fable 5.162.2
  3. #9Kimi K351.9
  4. ~#10Phi-4-reasoning-plusEstimated48.0
  5. #10Grok 4.647.0
No published score — estimated from related benchmarks

Hard reasoning

AA Intelligence IndexEstimated
~#50 of 75 (estimated)~16.3
  1. #1Claude Opus 5.557.6
  2. #2Claude Sonnet 5.556.0
  3. #49Ring-2.6-1T16.6
  4. ~#50Phi-4-reasoning-plusEstimated16.3
  5. #50K2 Horizon 3.7B15.6
No published score — estimated from related benchmarks