Microsoft Researchopen-weight
Released: 2024-12-12

Phi-4

Microsoft 14B state-of-the-art small language model trained with high-quality synthetic data for complex reasoning.

Intelligence Index
75.0/ 100
Calibrated multi-domain composite
Throughput Speed
85tok/sec
Output streaming throughput
Time To First Token
260ms
Initial chunk server latency
LiveBench Contam-Free
—%
September 2026 suite score

Sourced Benchmark Evaluations (2)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
ARC-AGI-2Nearing Saturation59.3%
75
2025-01-09independentSource ↗
SWE-bench VerifiedSaturated47.1%
25
2024-12-21independentSource ↗
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run phi-4

# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "phi-4",
    "messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
  }'

Evaluation & Sourcing Notes

Scores listed for Phi-4 represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.