Model Profile
Phi-3-medium-128k-instruct
Use this page to decide where this model is a strong fit. Rankings below are benchmark-backed by use case, with explicit confidence and contributor metrics.
Identity
ID: microsoft/Phi-3-medium-128k-instruct
Author: microsoft
Origin: huggingface_catalog
Arch: unknown
Benchmark Coverage
Scored use cases: 12
Avg confidence: 15.9%
Evidence points: 56
Raw rows: 94
Weighted rows: 7
Catalog Metadata
Parameters: unknown
Context window: 4096
Downloads: 52,521
Intelligence Profile
Dimension Breakdown
No eq benchmarks found
No creativity benchmarks found
No based benchmarks found
* Low confidence — limited benchmark evidence for this dimension
2/5 dimensions scored · Last updated Apr 2, 2026
Benchmark Signals
Click through to the benchmark source behind this model profile.
DuckDB NSQL Leaderboard
all_execution_accuracy
Normalized value 55.8% · confidence 100.0%
Strongest impact in Metric definition workshop
duckdb_nsql_leaderboard.all_execution_accuracy · Apr 1, 2026
RepoQA Official Results
overall_average_pass_at_1_pct
Normalized value 74.4% · confidence 100.0%
Strongest impact in Debugging assistant
repoqa_leaderboard.overall_average_pass_at_1_pct · Apr 1, 2026
DuckDB NSQL Leaderboard
hard_execution_accuracy
Normalized value 50.0% · confidence 100.0%
Strongest impact in SQL debugging
duckdb_nsql_leaderboard.hard_execution_accuracy · Apr 1, 2026
BigCodeBench Official
bigcodebench_complete_pct
Normalized value 74.9% · confidence 100.0%
Strongest impact in Verilog/VHDL generation
bigcodebench_official.bigcodebench_complete_pct · Apr 1, 2026
RepoQA Official Results
all_average_pass_at_1_pct
Normalized value 74.4% · confidence 100.0%
Strongest impact in Unit test generation
repoqa_leaderboard.all_average_pass_at_1_pct · Apr 1, 2026
BigCodeBench Official
bigcodebench_instruct_pct
Normalized value 69.4% · confidence 100.0%
Strongest impact in Integration test generation
bigcodebench_official.bigcodebench_instruct_pct · Apr 1, 2026
Some fit rows have limited benchmark evidence.
12 of 12 scored use cases have low confidence or thin contributor coverage.
Coverage Diagnostics
actively scoredUse-Case Scores
17
Total Measurements
94
Weighted Measurements
7
Weighted Sources
3
Raw Source Coverage
Weighted Source Coverage
Best Use Cases for This Model
| Use Case | Score |
|---|---|
| Debugging assistant use_case.dev.debugging | 13.3% |
| Unit test generation use_case.dev.test_generation | 12.3% |
| Code Review Assistant use_case.dev.code_review_assistant | 11.5% |
| Integration test generation use_case.dev.integration_tests | 10.7% |
| Metric definition workshop use_case.data.metric_definition_workshop | 10.4% |
| Refactoring assistant use_case.dev.refactoring | 10.3% |
| Verilog/VHDL generation use_case.eda.verilog_generation | 9.8% |
| SQL debugging use_case.data.sql_debugging | 9.4% |
| Documentation from code use_case.dev.docstrings_and_docs | 8.7% |
| Data quality assistant use_case.data.data_quality_assistant | 8.2% |
| Text-to-SQL analyst assistant use_case.data.text_to_sql | 7.5% |
| Executive brief from metrics use_case.data.exec_brief_from_metrics | 7.4% |