Table of Contents

DuckDB.NET 1.5

This page lists the changes in DuckDB.NET 1.5 releases, newest first. For other versions, see Release Notes.

1.5.6.1

Released 2026-10-15 (NuGet, changes since 1.5.6). Built on DuckDB v1.5.6.

New features

  • Read a TIME column as TimeSpan.
  • Read FLOAT as double, DOUBLE as float, and both as decimal, in columns, list items, maps and structs.
  • Read HUGEINT and UHUGEINT columns as DuckDBHugeInt and DuckDBUHugeInt with GetFieldValue and GetProviderSpecificValue.
  • The mapped appender now accepts DuckDBDateOnly and DuckDBTimeOnly properties, and scalar functions accept them as parameters and results.
  • BigInteger values can now be bound and appended to UHUGEINT columns as unsigned values, including through the mapped appender.
  • A table function over an array now uses the length of the array as its cardinality hint when none is given.

Performance

These numbers were measured against 1.5.6 on .NET 10 with a million rows. The gains are in moving values between DuckDB and .NET, so bulk appends, reading large results and queries that call C# functions benefit most. Queries dominated by DuckDB's own work change little.

  • Reading numbers and booleans through GetInt32, GetFieldValue<T> and the other typed getters is 3.5 to 6 times faster. Reading an integer column as double is 12 times faster.
  • Reading short strings is up to 1.35 times faster.
  • Reading dates, times and timestamps is 3 to 12 times faster. Appending a DateTime is 14 to 26 times faster.
  • UUID, DECIMAL and HUGEINT reads are 1.7 to 7 times faster.
  • Typed scalar functions are 5 to 8 times faster, and 1.6 times faster with a string argument.
  • Table functions are 2.5 to 6 times faster.
  • With projection pushdown, the columns a query does not read are no longer computed. A query that reads one column of a four-column table function is about 6 times faster, and the gain grows with the number of unread columns.
  • LIST, STRUCT and MAP columns are read into .NET types up to 6 times faster.
  • Appending BigInteger, lists, strings, C# enums and through the mapped appender is 1.3 to 4 times faster. Numbers are 1.2 to 1.4 times faster. Decimals of up to 18 digits are 1.3 to 2 times faster depending on the CPU, and wider decimals more than twice as fast.

Memory

These numbers are per million rows, against 1.5.6.

  • Appending no longer allocates for each value. 1.5.6 allocated about 750 bytes for every DateTime (2.9 GB for four TIMESTAMP columns), 756 MB for a record of mixed types in the mapped appender, 211 MB for lists of int?, 188 MB for C# enums and up to 315 MB for BigInteger. Each is now about 0.1 MB.
  • Scalar functions with nullable arguments or results, and table functions over structs or rows with nullable columns, no longer allocate for each row. They allocated 21 to 47 MB in 1.5.6 and now about 0.1 MB.
  • An integer read as double no longer allocates (375 MB for eight million values in 1.5.6). A DECIMAL wider than 18 digits is read without a BigInteger (107 MB, now none). A large HUGEINT read as BigInteger allocates a third as much.
  • LIST, STRUCT and MAP reads allocate up to 60% less, and GetValues 9 to 17% less.

Fixes

  • A Read() after the end of a streaming result, and the first Read() of an empty streaming result, no longer throw (#358).
  • Reading an ENUM column as a type other than string no longer reads past the stored value.
  • Reading a BIT column as an unsupported type throws InvalidCastException instead of ending the process with a stack overflow.
  • Negative intervals can be read as TimeSpan, bound and appended.
  • A TIMETZ read as DateTimeOffset no longer throws for a time earlier in the day than its positive offset.
  • -2^127 is accepted as a HUGEINT value, and the HUGEINT limits no longer fail under cultures with another minus sign.
  • A decimal parameter is sent with the exact number of its digits. Before, 1000 was sent as DECIMAL(3,0).
  • A list parameter that DuckDB cannot build is refused instead of crashing the process.
  • The exception of a scalar function that throws in WHERE or HAVING is now the InnerException of the error.
  • A table function projection that calls a constructor (a record or a tuple) is refused with a clear message instead of a NullReferenceException.
  • The mapped appender disposes its appender when validation fails.
  • DuckDBClientFactory.CanCreateCommandBuilder and CanCreateDataAdapter return false instead of throwing.

Behavior changes

  • Appending a decimal too wide for its column throws instead of storing a value the column cannot hold.
  • A TimeSpan parameter where DuckDB cannot infer the type is bound as an INTERVAL instead of as text. Declare DbType.Time to bind it as a time.
  • Looking up a missing parameter by name through DbParameterCollection throws IndexOutOfRangeException, as the typed indexer does.
  • A TIMETZ read as DateTimeOffset is placed on 0001-01-02 instead of 0001-01-01.
  • A BigInteger parameter where DuckDB cannot infer the type is bound as HUGEINT if it fits, as UHUGEINT up to 2^128 - 1 and as BIGNUM beyond. Before, values outside the HUGEINT range threw.

Bindings

  • New features:
    • DuckDBHugeInt.ToInt128 and DuckDBUHugeInt.ToUInt128.
    • Equality without boxing (IEquatable<T>) for DuckDBDateOnly, DuckDBTimeOnly, DuckDBTimestamp, DuckDBInterval and DuckDBTimeTz.
    • Bindings for duckdb_scalar_function_set_bind_data_copy and duckdb_unsafe_vector_assign_string_element_len.
  • Performance: creating a DuckDBHugeInt or DuckDBUHugeInt from a BigInteger is faster and no longer allocates.
  • Fixes:
    • DuckDBInterval reads and writes negative days and microseconds.
    • DuckDBHugeInt accepts -2^127, and its limits no longer fail under cultures with another minus sign.

1.5.6

Released 2026-09-28 (NuGet, changes since 1.5.5). Built on DuckDB v1.5.6.

New features

  • Typed properties for 24 DuckDB configuration options on DuckDBConnectionStringBuilder.
  • Read DECIMAL values that do not fit System.Decimal as BigInteger or DuckDBDecimal.
  • Bind BigInteger parameters wider than 128 bits to BIGNUM (formerly VARINT) columns.

Fixes

  • Fixed possible data loss on Linux and macOS when a reader opened with CommandBehavior.CloseConnection closed the last connection to a file database (#357).
  • Errors in later chunks of a streaming result are now thrown instead of silently ending the result, in DuckDBDataReader and the Arrow stream (#356).
  • Fixed a memory leak where results skipped with NextResult were never released.
  • Fixed incomplete appender rows being written with uninitialized values (#355).
  • Fixed ARRAY values shifted by one row after a NULL array in the appender.
  • Fixed the appender writing to freed memory when a list of arrays grows.
  • Fixed table functions returning no rows after the first iteration of a recursive CTE (#348).

Behavior changes

  • GetFieldValue<object> now returns the same value as GetValue.
  • Reading a LIST or MAP into an untyped target (object, ArrayList, List<object>, or a Dictionary with object keys or values) throws InvalidCastException. Use GetValue instead.
  • A failed ExecuteReader with CommandBehavior.CloseConnection now closes the connection, as in Npgsql.
  • Moving on from an incomplete appender row discards it and faults the appender: the next CreateRow or AppendRow throws, and rows completed before it are still written. Close, Dispose and Clear discard an incomplete row without an error.
  • A blob Stream from GetStream is valid only until the reader moves to the next result set.

Bindings

  • Added a binding for duckdb_create_bignum and the DuckDBBignum struct.

1.5.5

Released 2026-07-26 (NuGet, changes since 1.5.3). Built on DuckDB v1.5.5.

New features

  • Apache Arrow result streaming through DuckDBCommand.ExecuteArrowStream and ExecuteArrowBatchesAsync (#26). See Apache Arrow.
  • A new allocation-free DuckDBAppender.AppendRow API with reusable rows.
  • Support for byte[] columns in DuckDBAppenderMap.

Fixes

  • Fixed IndexOutOfRangeException when accessing reader metadata after exhausting the result set (#332).

Performance

  • Reduced allocations in parameter binding and same-type object reads.

Bindings

  • Added bindings for the Arrow C Data Interface (duckdb_to_arrow_schema, duckdb_data_chunk_to_arrow, duckdb_result_get_arrow_options).
  • Added bindings for error data (duckdb_appender_error_data, duckdb_error_data_error_type).

1.5.3

Released 2026-05-24 (NuGet, changes since 1.5.2). Built on DuckDB v1.5.3.

New features

  • Improved support for binding ICollection parameters (#326).

Fixes

  • Fixed ObjectDisposedException in the enum appender (#327).

1.5.2

Released 2026-04-24 (NuGet, changes since 1.5.0). Built on DuckDB v1.5.2.

New features

  • Projection pushdown for table functions.
  • Estimated cardinality hints for table functions.
  • IList<T> column support in table functions.

1.5.0

Released 2026-03-10 (NuGet, changes since 1.4.4). Built on DuckDB v1.5.0.

New features

  • High-level scalar function API. See Scalar Functions.
  • High-level table function API. See Table Functions.
  • Named parameters for table functions through the [Named] attribute.
  • NULL handling for scalar functions is inferred from nullable parameter types (int?, string?).
  • An appender Clear() method to discard in-progress rows.

Performance

  • Vector readers are reused across chunks instead of being recreated for each chunk.
  • Migrated from DllImport to LibraryImport with source-generated P/Invoke.
  • Enum dictionary lookups in EnumVectorDataReader are cached.
  • Nullable GetValue<T> uses expression trees to avoid boxing.
  • SuppressGCTransition on trivial native calls.
  • AggressiveInlining on hot-path methods.
  • Appender value writes are split by struct and class to avoid boxing.

Behavior changes

  • Dropped netstandard2.0 and net6.0. The packages now target net8.0 and later.

Bindings

  • Custom string marshallers for correct handling of DuckDB string ownership.
Edit this page