bedrock: Add GPT-5.6 Sol, Terra, and Luna Mantle models (#61008)
# Objective - Adds the three GPT-5.6 models (Sol, Terra, and Luna) that OpenAI now serves through AWS Bedrock's `bedrock-mantle` endpoint, so they can be selected as built-in models under the Amazon Bedrock provider. - Follows up on the native Bedrock Mantle support added in #60480. - Addresses the feature request in https://github.com/zed-industries/zed/discussions/61003. ## Solution - Adds `Gpt5_6Sol`, `Gpt5_6Terra`, and `Gpt5_6Luna` variants to `MantleModel` in `crates/bedrock/src/models.rs`, with per-model `id`/`request_id`/`display_name` and shared capability arms. - All three are Responses-API, `bedrock-mantle`-only models, so they reuse the existing Mantle request/response plumbing, region gating, bearer-token auth, and model-picker wiring (`MantleModel::iter()`) with no other changes required. - Metadata mirrors the AWS Bedrock model cards ([Sol](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-sol.html), [Terra](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-terra.html), [Luna](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-56-luna.html)): 272K context window, image input, tool and thinking support, and `openai.gpt-5.6-{sol,terra,luna}` request IDs. - Updates the Amazon Bedrock section of "Use a Gateway" to mention the GPT-5.6 family. ## Testing - `cargo test -p bedrock` — adds `test_gpt_5_6_mantle_model_metadata` and extends `test_builtin_mantle_models_use_responses_protocol`; all pass. - `cargo test -p language_models mantle` — the consumer crate compiles and all Mantle tests pass. - `./script/clippy -p bedrock` passes with no new warnings; docs pass `prettier --check`. ## Self-Review Checklist: - [x] I've reviewed my own diff for quality, security, and reliability - [x] Unsafe blocks (if any) have justifying comments — N/A, no unsafe code - [x] The content adheres to Zed's UI standards — N/A, no UI change - [x] Tests cover the new/changed behavior - [x] Performance impact has been considered and is acceptable --- Release Notes: - Added GPT-5.6 Sol, Terra, and Luna models to the Amazon Bedrock provider via the `bedrock-mantle` endpoint. --------- Co-authored-by: Anant Goel <anant@zed.dev>
This commit is contained in:
Ibrahim KhanandAnant Goel authored and GitHub committed
1 parent
7862fff55c
commit
48ba4bada2
4 files changed
+140
-9
No files matched your search
@@ -834,6 +834,12 @@ pub enum MantleProtocol {
|
||||
#[cfg_attr(feature = "schemars", derive(schemars::JsonSchema))]
|
||||
#[derive(Clone, Debug, Serialize, Deserialize, PartialEq, EnumIter)]
|
||||
pub enum MantleModel {
|
||||
#[serde(rename = "gpt-5.6-sol")]
|
||||
Gpt5_6Sol,
|
||||
#[serde(rename = "gpt-5.6-terra")]
|
||||
Gpt5_6Terra,
|
||||
#[serde(rename = "gpt-5.6-luna")]
|
||||
Gpt5_6Luna,
|
||||
#[serde(rename = "gpt-5.5")]
|
||||
Gpt5_5,
|
||||
#[serde(rename = "gpt-5.4")]
|
||||
@@ -857,6 +863,9 @@ impl MantleModel {
|
||||
/// The model id Zed uses internally (also used as the `name` in settings).
|
||||
pub fn id(&self) -> &str {
|
||||
match self {
|
||||
Self::Gpt5_6Sol => "gpt-5.6-sol",
|
||||
Self::Gpt5_6Terra => "gpt-5.6-terra",
|
||||
Self::Gpt5_6Luna => "gpt-5.6-luna",
|
||||
Self::Gpt5_5 => "gpt-5.5",
|
||||
Self::Gpt5_4 => "gpt-5.4",
|
||||
Self::Grok4_3 => "grok-4.3",
|
||||
@@ -867,6 +876,9 @@ impl MantleModel {
|
||||
/// The model id as expected in Bedrock Mantle request bodies, e.g. `openai.gpt-5.5`.
|
||||
pub fn request_id(&self) -> &str {
|
||||
match self {
|
||||
Self::Gpt5_6Sol => "openai.gpt-5.6-sol",
|
||||
Self::Gpt5_6Terra => "openai.gpt-5.6-terra",
|
||||
Self::Gpt5_6Luna => "openai.gpt-5.6-luna",
|
||||
Self::Gpt5_5 => "openai.gpt-5.5",
|
||||
Self::Gpt5_4 => "openai.gpt-5.4",
|
||||
Self::Grok4_3 => "xai.grok-4.3",
|
||||
@@ -876,6 +888,9 @@ impl MantleModel {
|
||||
|
||||
pub fn display_name(&self) -> &str {
|
||||
match self {
|
||||
Self::Gpt5_6Sol => "GPT-5.6 Sol",
|
||||
Self::Gpt5_6Terra => "GPT-5.6 Terra",
|
||||
Self::Gpt5_6Luna => "GPT-5.6 Luna",
|
||||
Self::Gpt5_5 => "GPT-5.5",
|
||||
Self::Gpt5_4 => "GPT-5.4",
|
||||
Self::Grok4_3 => "Grok 4.3",
|
||||
@@ -888,14 +903,23 @@ impl MantleModel {
|
||||
/// Which OpenAI-compatible API this model must be called through.
|
||||
pub fn protocol(&self) -> MantleProtocol {
|
||||
match self {
|
||||
Self::Gpt5_5 | Self::Gpt5_4 | Self::Grok4_3 => MantleProtocol::Responses,
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4
|
||||
| Self::Grok4_3 => MantleProtocol::Responses,
|
||||
Self::Custom { protocol, .. } => *protocol,
|
||||
}
|
||||
}
|
||||
|
||||
pub fn max_token_count(&self) -> u64 {
|
||||
match self {
|
||||
Self::Gpt5_5 | Self::Gpt5_4 => 272_000,
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4 => 272_000,
|
||||
Self::Grok4_3 => 1_000_000,
|
||||
Self::Custom { max_tokens, .. } => *max_tokens,
|
||||
}
|
||||
@@ -903,8 +927,12 @@ impl MantleModel {
|
||||
|
||||
pub fn max_output_tokens(&self) -> u64 {
|
||||
match self {
|
||||
// AWS doesn't document a hard cap for GPT-5.5/5.4 on Mantle.
|
||||
Self::Gpt5_5 | Self::Gpt5_4 => 128_000,
|
||||
// AWS doesn't document a hard cap for the GPT-5.x models on Mantle.
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4 => 128_000,
|
||||
Self::Grok4_3 => 131_072,
|
||||
Self::Custom {
|
||||
max_output_tokens, ..
|
||||
@@ -914,14 +942,24 @@ impl MantleModel {
|
||||
|
||||
pub fn supports_tools(&self) -> bool {
|
||||
match self {
|
||||
Self::Gpt5_5 | Self::Gpt5_4 | Self::Grok4_3 => true,
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4
|
||||
| Self::Grok4_3 => true,
|
||||
Self::Custom { supports_tools, .. } => *supports_tools,
|
||||
}
|
||||
}
|
||||
|
||||
pub fn supports_images(&self) -> bool {
|
||||
match self {
|
||||
Self::Gpt5_5 | Self::Gpt5_4 | Self::Grok4_3 => true,
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4
|
||||
| Self::Grok4_3 => true,
|
||||
Self::Custom {
|
||||
supports_images, ..
|
||||
} => *supports_images,
|
||||
@@ -930,7 +968,12 @@ impl MantleModel {
|
||||
|
||||
pub fn supports_thinking(&self) -> bool {
|
||||
match self {
|
||||
Self::Gpt5_5 | Self::Gpt5_4 | Self::Grok4_3 => true,
|
||||
Self::Gpt5_6Sol
|
||||
| Self::Gpt5_6Terra
|
||||
| Self::Gpt5_6Luna
|
||||
| Self::Gpt5_5
|
||||
| Self::Gpt5_4
|
||||
| Self::Grok4_3 => true,
|
||||
Self::Custom {
|
||||
supports_thinking, ..
|
||||
} => *supports_thinking,
|
||||
@@ -944,11 +987,84 @@ mod tests {
|
||||
|
||||
#[test]
|
||||
fn test_builtin_mantle_models_use_responses_protocol() {
|
||||
assert_eq!(MantleModel::Gpt5_6Sol.protocol(), MantleProtocol::Responses);
|
||||
assert_eq!(
|
||||
MantleModel::Gpt5_6Terra.protocol(),
|
||||
MantleProtocol::Responses
|
||||
);
|
||||
assert_eq!(
|
||||
MantleModel::Gpt5_6Luna.protocol(),
|
||||
MantleProtocol::Responses
|
||||
);
|
||||
assert_eq!(MantleModel::Gpt5_5.protocol(), MantleProtocol::Responses);
|
||||
assert_eq!(MantleModel::Gpt5_4.protocol(), MantleProtocol::Responses);
|
||||
assert_eq!(MantleModel::Grok4_3.protocol(), MantleProtocol::Responses);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_gpt_5_6_mantle_model_metadata() {
|
||||
// Values mirror the AWS Bedrock model cards for GPT-5.6 Sol/Terra/Luna,
|
||||
// which are served only through the `bedrock-mantle` Responses API.
|
||||
for (model, id, request_id, display_name) in [
|
||||
(
|
||||
MantleModel::Gpt5_6Sol,
|
||||
"gpt-5.6-sol",
|
||||
"openai.gpt-5.6-sol",
|
||||
"GPT-5.6 Sol",
|
||||
),
|
||||
(
|
||||
MantleModel::Gpt5_6Terra,
|
||||
"gpt-5.6-terra",
|
||||
"openai.gpt-5.6-terra",
|
||||
"GPT-5.6 Terra",
|
||||
),
|
||||
(
|
||||
MantleModel::Gpt5_6Luna,
|
||||
"gpt-5.6-luna",
|
||||
"openai.gpt-5.6-luna",
|
||||
"GPT-5.6 Luna",
|
||||
),
|
||||
] {
|
||||
assert_eq!(model.id(), id);
|
||||
assert_eq!(model.request_id(), request_id);
|
||||
assert_eq!(model.display_name(), display_name);
|
||||
assert_eq!(model.max_token_count(), 272_000);
|
||||
assert!(model.supports_tools());
|
||||
assert!(model.supports_images());
|
||||
assert!(model.supports_thinking());
|
||||
}
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_builtin_mantle_models_have_unique_ids_and_sane_token_limits() {
|
||||
use std::collections::HashSet;
|
||||
use strum::IntoEnumIterator;
|
||||
|
||||
let mut ids = HashSet::new();
|
||||
let mut request_ids = HashSet::new();
|
||||
for model in
|
||||
MantleModel::iter().filter(|model| !matches!(model, MantleModel::Custom { .. }))
|
||||
{
|
||||
assert!(
|
||||
ids.insert(model.id().to_string()),
|
||||
"duplicate MantleModel id: {}",
|
||||
model.id()
|
||||
);
|
||||
assert!(
|
||||
request_ids.insert(model.request_id().to_string()),
|
||||
"duplicate MantleModel request_id: {}",
|
||||
model.request_id()
|
||||
);
|
||||
assert!(
|
||||
model.max_output_tokens() <= model.max_token_count(),
|
||||
"{} has max_output_tokens ({}) greater than max_token_count ({})",
|
||||
model.id(),
|
||||
model.max_output_tokens(),
|
||||
model.max_token_count()
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_us_region_inference_ids() -> anyhow::Result<()> {
|
||||
assert_eq!(
|
||||
|
||||
@@ -2606,9 +2606,24 @@ mod tests {
|
||||
|
||||
#[test]
|
||||
fn test_builtin_mantle_models_support_thinking() {
|
||||
assert!(MantleModel::Gpt5_6Sol.supports_thinking());
|
||||
assert!(MantleModel::Gpt5_6Terra.supports_thinking());
|
||||
assert!(MantleModel::Gpt5_6Luna.supports_thinking());
|
||||
assert!(MantleModel::Gpt5_5.supports_thinking());
|
||||
assert!(MantleModel::Gpt5_4.supports_thinking());
|
||||
assert!(MantleModel::Grok4_3.supports_thinking());
|
||||
assert_eq!(
|
||||
mantle_default_reasoning_effort(&MantleModel::Gpt5_6Sol),
|
||||
Some(ReasoningEffort::Medium)
|
||||
);
|
||||
assert_eq!(
|
||||
mantle_default_reasoning_effort(&MantleModel::Gpt5_6Terra),
|
||||
Some(ReasoningEffort::Medium)
|
||||
);
|
||||
assert_eq!(
|
||||
mantle_default_reasoning_effort(&MantleModel::Gpt5_6Luna),
|
||||
Some(ReasoningEffort::Medium)
|
||||
);
|
||||
assert_eq!(
|
||||
mantle_default_reasoning_effort(&MantleModel::Gpt5_5),
|
||||
Some(ReasoningEffort::Medium)
|
||||
|
||||
@@ -118,7 +118,7 @@ pub struct AmazonBedrockSettingsContent {
|
||||
pub available_models: Option<Vec<BedrockAvailableModel>>,
|
||||
/// Custom models served through the `bedrock-mantle` endpoint's
|
||||
/// OpenAI-compatible APIs, in addition to the built-in Mantle models
|
||||
/// (GPT-5.5, GPT-5.4, Grok 4.3).
|
||||
/// (GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, Grok 4.3).
|
||||
pub mantle_available_models: Option<Vec<BedrockMantleAvailableModel>>,
|
||||
pub custom_headers: Option<HashMap<String, String>>,
|
||||
pub endpoint_url: Option<String>,
|
||||
|
||||
@@ -215,7 +215,7 @@ Some AWS environments require a guardrail on every Bedrock API call. Add `guardr
|
||||
|
||||
### Bedrock Mantle Models {#bedrock-mantle-models}
|
||||
|
||||
Some models, such as GPT-5.5, GPT-5.4, and Grok 4.3, aren't available through Bedrock's Converse API and are only reachable through `bedrock-mantle`, AWS's OpenAI-compatible inference endpoint. Zed routes these models through `bedrock-mantle` automatically; they appear alongside the rest of the Bedrock models in the model picker once you're authenticated, with no extra configuration required.
|
||||
Some models, such as the GPT-5.6 family (Sol, Terra, and Luna), GPT-5.5, GPT-5.4, and Grok 4.3, aren't available through Bedrock's Converse API and are only reachable through `bedrock-mantle`, AWS's OpenAI-compatible inference endpoint. Zed routes these models through `bedrock-mantle` automatically; they appear alongside the rest of the Bedrock models in the model picker once you're authenticated, with no extra configuration required.
|
||||
|
||||
Mantle models require IAM permissions for the `bedrock-mantle` endpoint (for example via the `AmazonBedrockMantleInferenceAccess` managed policy) in addition to whatever permissions your existing Bedrock credentials already have, and `bedrock-mantle` is only available in [some AWS Regions](https://docs.aws.amazon.com/bedrock/latest/userguide/bedrock-mantle.html#regions). Zed surfaces an error naming the current Region and the supported ones if you try to use a Mantle model outside of them.
|
||||
|
||||
|
||||
Reference in new issue
Block a user