diff --git a/meilitool/src/upgrade/mod.rs b/meilitool/src/upgrade/mod.rs
deleted file mode 100644
index ae095b6bd..000000000
--- a/meilitool/src/upgrade/mod.rs
+++ /dev/null
@@ -1,73 +0,0 @@
-mod v1_10;
-mod v1_11;
-mod v1_9;
-
-use std::path::{Path, PathBuf};
-
-use anyhow::{bail, Context};
-use meilisearch_types::versioning::create_version_file;
-
-use v1_10::v1_9_to_v1_10;
-
-use crate::upgrade::v1_11::v1_10_to_v1_11;
-
-pub struct OfflineUpgrade {
-    pub db_path: PathBuf,
-    pub current_version: (String, String, String),
-    pub target_version: (String, String, String),
-}
-
-impl OfflineUpgrade {
-    pub fn upgrade(self) -> anyhow::Result<()> {
-        let upgrade_list = [
-            (v1_9_to_v1_10 as fn(&Path) -> Result<(), anyhow::Error>, "1", "10", "0"),
-            (v1_10_to_v1_11, "1", "11", "0"),
-        ];
-
-        let (current_major, current_minor, current_patch) = &self.current_version;
-
-        let start_at = match (
-            current_major.as_str(),
-            current_minor.as_str(),
-            current_patch.as_str(),
-        ) {
-            ("1", "9", _) => 0,
-            ("1", "10", _) => 1,
-            _ => {
-                bail!("Unsupported current version {current_major}.{current_minor}.{current_patch}. Can only upgrade from v1.9")
-            }
-        };
-
-        let (target_major, target_minor, target_patch) = &self.target_version;
-
-        let ends_at = match (target_major.as_str(), target_minor.as_str(), target_patch.as_str()) {
-            ("1", "10", _) => 0,
-            ("1", "11", _) => 1,
-            (major, _, _) if major.starts_with('v') => {
-                bail!("Target version must not starts with a `v`. Instead of writing `v1.9.0` write `1.9.0` for example.")
-            }
-            _ => {
-                bail!("Unsupported target version {target_major}.{target_minor}.{target_patch}. Can only upgrade to v1.11")
-            }
-        };
-
-        println!("Starting the upgrade from {current_major}.{current_minor}.{current_patch} to {target_major}.{target_minor}.{target_patch}");
-
-        #[allow(clippy::needless_range_loop)]
-        for index in start_at..=ends_at {
-            let (func, major, minor, patch) = upgrade_list[index];
-            (func)(&self.db_path)?;
-            println!("Done");
-            // We're writing the version file just in case an issue arise _while_ upgrading.
-            // We don't want the DB to fail in an unknown state.
-            println!("Writing VERSION file");
-
-            create_version_file(&self.db_path, major, minor, patch)
-                .context("while writing VERSION file after the upgrade")?;
-        }
-
-        println!("Success");
-
-        Ok(())
-    }
-}
diff --git a/meilitool/src/upgrade/v1_10.rs b/meilitool/src/upgrade/v1_10.rs
deleted file mode 100644
index 671f4d6d2..000000000
--- a/meilitool/src/upgrade/v1_10.rs
+++ /dev/null
@@ -1,289 +0,0 @@
-use anyhow::bail;
-use std::path::Path;
-
-use anyhow::Context;
-use meilisearch_types::{
-    heed::{
-        types::{SerdeJson, Str},
-        Database, Env, EnvOpenOptions, RoTxn, RwTxn, Unspecified,
-    },
-    milli::index::{db_name, main_key},
-};
-
-use crate::{try_opening_database, try_opening_poly_database, uuid_codec::UuidCodec};
-
-use super::v1_9;
-
-pub type FieldDistribution = std::collections::BTreeMap<String, u64>;
-
-/// The statistics that can be computed from an `Index` object.
-#[derive(serde::Serialize, serde::Deserialize, Debug)]
-pub struct IndexStats {
-    /// Number of documents in the index.
-    pub number_of_documents: u64,
-    /// Size taken up by the index' DB, in bytes.
-    ///
-    /// This includes the size taken by both the used and free pages of the DB, and as the free pages
-    /// are not returned to the disk after a deletion, this number is typically larger than
-    /// `used_database_size` that only includes the size of the used pages.
-    pub database_size: u64,
-    /// Size taken by the used pages of the index' DB, in bytes.
-    ///
-    /// As the DB backend does not return to the disk the pages that are not currently used by the DB,
-    /// this value is typically smaller than `database_size`.
-    pub used_database_size: u64,
-    /// Association of every field name with the number of times it occurs in the documents.
-    pub field_distribution: FieldDistribution,
-    /// Creation date of the index.
-    #[serde(with = "time::serde::rfc3339")]
-    pub created_at: time::OffsetDateTime,
-    /// Date of the last update of the index.
-    #[serde(with = "time::serde::rfc3339")]
-    pub updated_at: time::OffsetDateTime,
-}
-
-impl From<v1_9::IndexStats> for IndexStats {
-    fn from(
-        v1_9::IndexStats {
-            number_of_documents,
-            database_size,
-            used_database_size,
-            field_distribution,
-            created_at,
-            updated_at,
-        }: v1_9::IndexStats,
-    ) -> Self {
-        IndexStats {
-            number_of_documents,
-            database_size,
-            used_database_size,
-            field_distribution,
-            created_at: created_at.0,
-            updated_at: updated_at.0,
-        }
-    }
-}
-
-#[derive(serde::Serialize, serde::Deserialize)]
-#[serde(transparent)]
-pub struct OffsetDateTime(#[serde(with = "time::serde::rfc3339")] pub time::OffsetDateTime);
-
-fn update_index_stats(
-    index_stats: Database<UuidCodec, Unspecified>,
-    index_uid: &str,
-    index_uuid: uuid::Uuid,
-    sched_wtxn: &mut RwTxn,
-) -> anyhow::Result<()> {
-    let ctx = || format!("while updating index stats for index `{index_uid}`");
-
-    let stats: Option<&str> = index_stats
-        .remap_data_type::<Str>()
-        .get(sched_wtxn, &index_uuid)
-        .with_context(ctx)
-        .with_context(|| "While reading value")?;
-    dbg!(stats);
-
-    let stats: Option<v1_9::IndexStats> = index_stats
-        .remap_data_type::<SerdeJson<v1_9::IndexStats>>()
-        .get(sched_wtxn, &index_uuid)
-        .with_context(ctx)
-        .with_context(|| "While reading value")?;
-
-    if let Some(stats) = stats {
-        let stats: self::IndexStats = stats.into();
-
-        index_stats
-            .remap_data_type::<SerdeJson<self::IndexStats>>()
-            .put(sched_wtxn, &index_uuid, &stats)
-            .with_context(ctx)
-            .with_context(|| "While writing value")?;
-    }
-
-    Ok(())
-}
-
-fn update_date_format(
-    index_uid: &str,
-    index_env: &Env,
-    index_wtxn: &mut RwTxn,
-) -> anyhow::Result<()> {
-    let main = try_opening_poly_database(index_env, index_wtxn, db_name::MAIN)
-        .with_context(|| format!("while updating date format for index `{index_uid}`"))?;
-
-    date_round_trip(index_wtxn, index_uid, main, main_key::CREATED_AT_KEY)?;
-    date_round_trip(index_wtxn, index_uid, main, main_key::UPDATED_AT_KEY)?;
-
-    Ok(())
-}
-
-fn find_rest_embedders(
-    index_uid: &str,
-    index_env: &Env,
-    index_txn: &RoTxn,
-) -> anyhow::Result<Vec<String>> {
-    let main = try_opening_poly_database(index_env, index_txn, db_name::MAIN)
-        .with_context(|| format!("while checking REST embedders for index `{index_uid}`"))?;
-
-    let mut rest_embedders = vec![];
-
-    for config in main
-        .remap_types::<Str, SerdeJson<Vec<v1_9::IndexEmbeddingConfig>>>()
-        .get(index_txn, main_key::EMBEDDING_CONFIGS)?
-        .unwrap_or_default()
-    {
-        if let v1_9::EmbedderOptions::Rest(_) = config.config.embedder_options {
-            rest_embedders.push(config.name);
-        }
-    }
-
-    Ok(rest_embedders)
-}
-
-fn date_round_trip(
-    wtxn: &mut RwTxn,
-    index_uid: &str,
-    db: Database<Unspecified, Unspecified>,
-    key: &str,
-) -> anyhow::Result<()> {
-    let datetime =
-        db.remap_types::<Str, SerdeJson<v1_9::LegacyTime>>().get(wtxn, key).with_context(|| {
-            format!("could not read `{key}` while updating date format for index `{index_uid}`")
-        })?;
-
-    if let Some(datetime) = datetime {
-        db.remap_types::<Str, SerdeJson<self::OffsetDateTime>>()
-            .put(wtxn, key, &self::OffsetDateTime(datetime.0))
-            .with_context(|| {
-                format!(
-                    "could not write `{key}` while updating date format for index `{index_uid}`"
-                )
-            })?;
-    }
-
-    Ok(())
-}
-
-pub fn v1_9_to_v1_10(db_path: &Path) -> anyhow::Result<()> {
-    println!("Upgrading from v1.9.0 to v1.10.0");
-    // 2 changes here
-
-    // 1. date format. needs to be done before opening the Index
-    // 2. REST embedders. We don't support this case right now, so bail
-
-    let index_scheduler_path = db_path.join("tasks");
-    let env = unsafe { EnvOpenOptions::new().max_dbs(100).open(&index_scheduler_path) }
-        .with_context(|| format!("While trying to open {:?}", index_scheduler_path.display()))?;
-
-    let mut sched_wtxn = env.write_txn()?;
-
-    let index_mapping: Database<Str, UuidCodec> =
-        try_opening_database(&env, &sched_wtxn, "index-mapping")?;
-
-    let index_stats: Database<UuidCodec, Unspecified> =
-        try_opening_database(&env, &sched_wtxn, "index-stats").with_context(|| {
-            format!("While trying to open {:?}", index_scheduler_path.display())
-        })?;
-
-    let index_count =
-        index_mapping.len(&sched_wtxn).context("while reading the number of indexes")?;
-
-    // FIXME: not ideal, we have to pre-populate all indexes to prevent double borrow of sched_wtxn
-    // 1. immutably for the iteration
-    // 2. mutably for updating index stats
-    let indexes: Vec<_> = index_mapping
-        .iter(&sched_wtxn)?
-        .map(|res| res.map(|(uid, uuid)| (uid.to_owned(), uuid)))
-        .collect();
-
-    let mut rest_embedders = Vec::new();
-
-    let mut unwrapped_indexes = Vec::new();
-
-    // check that update can take place
-    for (index_index, result) in indexes.into_iter().enumerate() {
-        let (uid, uuid) = result?;
-        let index_path = db_path.join("indexes").join(uuid.to_string());
-
-        println!(
-            "[{}/{index_count}]Checking that update can take place for  `{uid}` at `{}`",
-            index_index + 1,
-            index_path.display()
-        );
-
-        let index_env = unsafe {
-            // FIXME: fetch the 25 magic number from the index file
-            EnvOpenOptions::new().max_dbs(25).open(&index_path).with_context(|| {
-                format!("while opening index {uid} at '{}'", index_path.display())
-            })?
-        };
-
-        let index_txn = index_env.read_txn().with_context(|| {
-            format!(
-                "while obtaining a write transaction for index {uid} at {}",
-                index_path.display()
-            )
-        })?;
-
-        println!("\t- Checking for incompatible embedders (REST embedders)");
-        let rest_embedders_for_index = find_rest_embedders(&uid, &index_env, &index_txn)?;
-
-        if rest_embedders_for_index.is_empty() {
-            unwrapped_indexes.push((uid, uuid));
-        } else {
-            // no need to add to unwrapped indexes because we'll exit early
-            rest_embedders.push((uid, rest_embedders_for_index));
-        }
-    }
-
-    if !rest_embedders.is_empty() {
-        let rest_embedders = rest_embedders
-            .into_iter()
-            .flat_map(|(index, embedders)| std::iter::repeat(index.clone()).zip(embedders))
-            .map(|(index, embedder)| format!("\t- embedder `{embedder}` in index `{index}`"))
-            .collect::<Vec<_>>()
-            .join("\n");
-        bail!("The update cannot take place because there are REST embedder(s). Remove them before proceeding with the update:\n{rest_embedders}\n\n\
-            The database has not been modified and is still a valid v1.9 database.");
-    }
-
-    println!("Update can take place, updating");
-
-    for (index_index, (uid, uuid)) in unwrapped_indexes.into_iter().enumerate() {
-        let index_path = db_path.join("indexes").join(uuid.to_string());
-
-        println!(
-            "[{}/{index_count}]Updating index `{uid}` at `{}`",
-            index_index + 1,
-            index_path.display()
-        );
-
-        let index_env = unsafe {
-            // FIXME: fetch the 25 magic number from the index file
-            EnvOpenOptions::new().max_dbs(25).open(&index_path).with_context(|| {
-                format!("while opening index {uid} at '{}'", index_path.display())
-            })?
-        };
-
-        let mut index_wtxn = index_env.write_txn().with_context(|| {
-            format!(
-                "while obtaining a write transaction for index `{uid}` at `{}`",
-                index_path.display()
-            )
-        })?;
-
-        println!("\t- Updating index stats");
-        update_index_stats(index_stats, &uid, uuid, &mut sched_wtxn)?;
-        println!("\t- Updating date format");
-        update_date_format(&uid, &index_env, &mut index_wtxn)?;
-
-        index_wtxn.commit().with_context(|| {
-            format!("while committing the write txn for index `{uid}` at {}", index_path.display())
-        })?;
-    }
-
-    sched_wtxn.commit().context("while committing the write txn for the index-scheduler")?;
-
-    println!("Upgrading database succeeded");
-
-    Ok(())
-}
diff --git a/meilitool/src/upgrade/v1_11.rs b/meilitool/src/upgrade/v1_11.rs
deleted file mode 100644
index 4105879fd..000000000
--- a/meilitool/src/upgrade/v1_11.rs
+++ /dev/null
@@ -1,86 +0,0 @@
-//! The breaking changes that happened between the v1.10 and the v1.11 are:
-//! - Arroy went from the v0.4.0 to the v0.5.0, see this release note to get the whole context: https://github.com/meilisearch/arroy/releases/tag/v0.5.0
-//!   - The `angular` distance has been renamed to `cosine` => We only need to update the string in the metadata.
-//!   - Reorganize the `NodeId` to make the appending of vectors work => We'll have to update the keys of almost all items in the DB.
-//!   - Store the list of updated IDs directly in LMDB instead of a roaring bitmap => This shouldn't be an issue since we are never supposed to commit this roaring bitmap, but it's not forbidden by arroy so ensuring it works is probably better than anything.
-
-use std::path::Path;
-
-use anyhow::Context;
-use meilisearch_types::{
-    heed::{types::Str, Database, EnvOpenOptions},
-    milli::index::db_name,
-};
-
-use crate::{try_opening_database, try_opening_poly_database, uuid_codec::UuidCodec};
-
-pub fn v1_10_to_v1_11(db_path: &Path) -> anyhow::Result<()> {
-    println!("Upgrading from v1.10.0 to v1.11.0");
-
-    let index_scheduler_path = db_path.join("tasks");
-    let env = unsafe { EnvOpenOptions::new().max_dbs(100).open(&index_scheduler_path) }
-        .with_context(|| format!("While trying to open {:?}", index_scheduler_path.display()))?;
-
-    let sched_rtxn = env.read_txn()?;
-
-    let index_mapping: Database<Str, UuidCodec> =
-        try_opening_database(&env, &sched_rtxn, "index-mapping")?;
-
-    let index_count =
-        index_mapping.len(&sched_rtxn).context("while reading the number of indexes")?;
-
-    let indexes: Vec<_> = index_mapping
-        .iter(&sched_rtxn)?
-        .map(|res| res.map(|(uid, uuid)| (uid.to_owned(), uuid)))
-        .collect();
-
-    // check that update can take place
-    for (index_index, result) in indexes.into_iter().enumerate() {
-        let (uid, uuid) = result?;
-        let index_path = db_path.join("indexes").join(uuid.to_string());
-
-        println!(
-            "[{}/{index_count}]Checking that update can take place for  `{uid}` at `{}`",
-            index_index + 1,
-            index_path.display()
-        );
-
-        let index_env = unsafe {
-            EnvOpenOptions::new().max_dbs(25).open(&index_path).with_context(|| {
-                format!("while opening index {uid} at '{}'", index_path.display())
-            })?
-        };
-
-        let index_rtxn = index_env.read_txn().with_context(|| {
-            format!(
-                "while obtaining a read transaction for index {uid} at {}",
-                index_path.display()
-            )
-        })?;
-        let index_read_database =
-            try_opening_poly_database(&index_env, &index_rtxn, db_name::VECTOR_ARROY)
-                .with_context(|| format!("while updating date format for index `{uid}`"))?;
-
-        let mut index_wtxn = index_env.write_txn().with_context(|| {
-            format!(
-                "while obtaining a write transaction for index {uid} at {}",
-                index_path.display()
-            )
-        })?;
-
-        let index_write_database =
-            try_opening_poly_database(&index_env, &index_wtxn, db_name::VECTOR_ARROY)
-                .with_context(|| format!("while updating date format for index `{uid}`"))?;
-
-        arroy_v04_to_v05::ugrade_from_prev_version(
-            &index_rtxn,
-            index_read_database,
-            &mut index_wtxn,
-            index_write_database,
-        )?;
-
-        index_wtxn.commit()?;
-    }
-
-    Ok(())
-}
diff --git a/meilitool/src/upgrade/v1_9.rs b/meilitool/src/upgrade/v1_9.rs
deleted file mode 100644
index 3e6cfde6c..000000000
--- a/meilitool/src/upgrade/v1_9.rs
+++ /dev/null
@@ -1,106 +0,0 @@
-use serde::{Deserialize, Serialize};
-use time::OffsetDateTime;
-
-pub type FieldDistribution = std::collections::BTreeMap<String, u64>;
-
-/// The statistics that can be computed from an `Index` object.
-#[derive(serde::Serialize, serde::Deserialize, Debug)]
-pub struct IndexStats {
-    /// Number of documents in the index.
-    pub number_of_documents: u64,
-    /// Size taken up by the index' DB, in bytes.
-    ///
-    /// This includes the size taken by both the used and free pages of the DB, and as the free pages
-    /// are not returned to the disk after a deletion, this number is typically larger than
-    /// `used_database_size` that only includes the size of the used pages.
-    pub database_size: u64,
-    /// Size taken by the used pages of the index' DB, in bytes.
-    ///
-    /// As the DB backend does not return to the disk the pages that are not currently used by the DB,
-    /// this value is typically smaller than `database_size`.
-    pub used_database_size: u64,
-    /// Association of every field name with the number of times it occurs in the documents.
-    pub field_distribution: FieldDistribution,
-    /// Creation date of the index.
-    pub created_at: LegacyTime,
-    /// Date of the last update of the index.
-    pub updated_at: LegacyTime,
-}
-
-#[derive(Debug, Deserialize, Serialize)]
-pub struct IndexEmbeddingConfig {
-    pub name: String,
-    pub config: EmbeddingConfig,
-}
-
-#[derive(Debug, Clone, Default, serde::Deserialize, serde::Serialize)]
-pub struct EmbeddingConfig {
-    /// Options of the embedder, specific to each kind of embedder
-    pub embedder_options: EmbedderOptions,
-}
-
-/// Options of an embedder, specific to each kind of embedder.
-#[derive(Debug, Clone, Hash, PartialEq, Eq, serde::Deserialize, serde::Serialize)]
-pub enum EmbedderOptions {
-    HuggingFace(hf::EmbedderOptions),
-    OpenAi(openai::EmbedderOptions),
-    Ollama(ollama::EmbedderOptions),
-    UserProvided(manual::EmbedderOptions),
-    Rest(rest::EmbedderOptions),
-}
-
-impl Default for EmbedderOptions {
-    fn default() -> Self {
-        Self::OpenAi(openai::EmbedderOptions { api_key: None, dimensions: None })
-    }
-}
-
-mod hf {
-    #[derive(Debug, Clone, Hash, PartialEq, Eq, serde::Deserialize, serde::Serialize)]
-    pub struct EmbedderOptions {
-        pub model: String,
-        pub revision: Option<String>,
-    }
-}
-mod openai {
-
-    #[derive(Debug, Clone, Hash, PartialEq, Eq, serde::Deserialize, serde::Serialize)]
-    pub struct EmbedderOptions {
-        pub api_key: Option<String>,
-        pub dimensions: Option<usize>,
-    }
-}
-mod ollama {
-    #[derive(Debug, Clone, Hash, PartialEq, Eq, serde::Deserialize, serde::Serialize)]
-    pub struct EmbedderOptions {
-        pub embedding_model: String,
-        pub url: Option<String>,
-        pub api_key: Option<String>,
-    }
-}
-mod manual {
-    #[derive(Debug, Clone, Hash, PartialEq, Eq, serde::Deserialize, serde::Serialize)]
-    pub struct EmbedderOptions {
-        pub dimensions: usize,
-    }
-}
-mod rest {
-    #[derive(Debug, Clone, PartialEq, Eq, serde::Deserialize, serde::Serialize, Hash)]
-    pub struct EmbedderOptions {
-        pub api_key: Option<String>,
-        pub dimensions: Option<usize>,
-        pub url: String,
-        pub input_field: Vec<String>,
-        // path to the array of embeddings
-        pub path_to_embeddings: Vec<String>,
-        // shape of a single embedding
-        pub embedding_object: Vec<String>,
-    }
-}
-
-// 2024-11-04 13:32:08.48368 +00:00:00
-time::serde::format_description!(legacy_datetime, OffsetDateTime, "[year]-[month]-[day] [hour]:[minute]:[second].[subsecond] [offset_hour sign:mandatory]:[offset_minute]:[offset_second]");
-
-#[derive(Debug, serde::Serialize, serde::Deserialize)]
-#[serde(transparent)]
-pub struct LegacyTime(#[serde(with = "legacy_datetime")] pub OffsetDateTime);