jerry/buzz
1
0
Fork 0
mirror of https://github.com/block/buzz.git synced 2026-10-27 04:30:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(agent): route GPT-5+ model-service FQNs to Responses (#7358)

Browse Source 
## Summary
Route Databricks Unity Catalog model services to OpenAI Responses when
their service name matches GPT-5 or newer. These models can reject tools
plus reasoning on Chat Completions.

Match only the service component, using the existing family-token
boundaries and a numeric major version. Catalog and schema names cannot
select the protocol. Keep neutral effort capabilities and the full model
ID unchanged; other services still use MLflow Chat Completions.

Keep the Rust and desktop resolvers in sync, add shared boundary cases
and a captured-HTTP regression for completion and summarization, and
update the documented FQN rule.

### Related issue
No duplicate found in searches for “FQN responses” PRs or “astra”
issues. Related: #6918 introduced Unity Catalog discovery.

Originating conversation:
buzz://message?channel=0b881928-a3a6-4c01-b981-8e64268f01ce&id=770949343bc96a9ed88acd90a1b37d358a0efc52c79237d0fdb491ce02b8d4ed

### Testing
No live Databricks inference test. The gateway must accept the full
model-service ID on its OpenAI Responses route; this remains the
integration risk.

The local `just ci` attempt exceeded its five-minute deadline during
`mobile-check`, so the full repository gate was not completed. All
push-hook checks passed.

Generated with Codex

Signed-off-by: Fizz <400e8babadcee6a7f420103f10a2849d84c4a9c71d5bd04f3948c814216648a3@buzz.block.builderlab.xyz>
Co-authored-by: Fizz <400e8babadcee6a7f420103f10a2849d84c4a9c71d5bd04f3948c814216648a3@buzz.block.builderlab.xyz>
This commit is contained in:
Bradley Axen
2026-09-08 07:23:48 -07:00
committed by GitHub
parent 86c189e857
commit cd54e26823
9 changed files with 559 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
crates/buzz-agent/README.md
View File

@@ -242,7 +242,7 @@ lifecycle hook — see [MCP_DRIVEN_HOOKS.md](../../docs/MCP_DRIVEN_HOOKS.md).
| Block Gateway | `openai` | `POST {base}/chat/completions` | gpt-5, claude |
| OpenRouter | `openrouter` | `POST {base}/chat/completions` | anything they route (extended-thinking replay, provider-agnostic tool calling) |
| Databricks | `databricks` | `POST {host}/serving-endpoints/{model}/invocations` | goose-claude-4-6-sonnet |
| Databricks AI Gateway v2 | `databricks_v2` | `POST {host}/ai-gateway/{provider}/v1/...` | workspace endpoints and Unity Catalog model-service FQNs; UC FQNs use MLflow Chat Completions |
| Databricks AI Gateway v2 | `databricks_v2` | `POST {host}/ai-gateway/{provider}/v1/...` | workspace endpoints and Unity Catalog model-service FQNs; UC GPT-5+ services use OpenAI Responses; other UC FQNs use MLflow Chat Completions |
The optional `DATABRICKS_MODEL_FILTER` applies only to model discovery. Each comma-separated entry is trimmed and matched against the complete raw ID with case-sensitive `*` (zero or more characters) and `?` (one Unicode character) semantics; patterns are OR-ed. Unset or blank preserves the full authenticated catalog. A nonblank value containing no usable patterns is rejected. This controls picker visibility only; Databricks and Unity Catalog permissions remain the authorization boundary. A filtered-empty result is authoritative and does not restore the built-in fallback models.

1
crates/buzz-agent/src/llm.rs
View File

@@ -2570,6 +2570,7 @@ fn apply_anthropic_cache_control(body: &mut serde_json::Map<String, Value>) {
#[cfg(test)]
mod tests {
include!("llm_fqn_tests.rs");
use super::*;
use crate::config::{Config, HookServers, OpenAiApi, Provider, ThinkingSummary};
use crate::types::{HistoryItem, ToolCall, ToolResult, ToolResultContent};

56
crates/buzz-agent/src/llm_fqn_tests.rs Normal file
View File

@@ -0,0 +1,56 @@
// Included in llm::tests to reuse the production-path HTTP capture fixture.
#[tokio::test]
async fn gpt_fqn_completion_and_summary_use_responses() {
let response = json!({"status":"completed", "output":[{
"type":"message", "content":[{"type":"output_text", "text":"ok"}]
}]});
let (base_url, captured) = spawn_sequence_stub(vec![
StubHttpResponse::ok(response.clone()),
StubHttpResponse::ok(response),
])
.await;
let mut config = cfg(Provider::DatabricksV2);
config.base_url = base_url;
config.thinking_effort = Some(ThinkingEffort::High);
let model = "catalog.schema.goose-gpt-6-astra";
let llm = Llm::new(&config).unwrap();
let tools = vec![ToolDef {
name: "test_tool".into(),
description: "Test".into(),
input_schema: json!({"type":"object", "properties":{}}),
}];
let result = llm
.complete(
&config,
"system",
&[HistoryItem::User("hello".into())],
&tools,
model,
)
.await
.unwrap();
assert_eq!(result.text, "ok");
assert_eq!(
llm.summarize(&config, "system", "history", 128, model)
.await
.unwrap(),
"ok"
);
let requests = captured.lock().await;
let posts: Vec<_> = requests.iter().filter(|r| r.method == "POST").collect();
assert_eq!(posts.len(), 2);
for request in &posts {
assert_eq!(request.path, "/v1/ai-gateway/openai/v1/responses");
let body = request.body.as_ref().unwrap();
assert_eq!(body["model"], model);
assert!(body.get("input").is_some());
assert!(body.get("messages").is_none());
assert!(body.get("reasoning_effort").is_none());
}
let completion = posts[0].body.as_ref().unwrap();
assert!(completion["input"].is_array());
assert_eq!(completion["reasoning"]["effort"], "high");
assert_eq!(completion["tools"][0]["type"], "function");
assert_eq!(completion["tools"][0]["name"], "test_tool");
assert_eq!(posts[1].body.as_ref().unwrap()["max_output_tokens"], 128);
}

60
crates/buzz-agent/src/model_capabilities.rs
View File

@@ -292,8 +292,8 @@ fn prefix_matches(token: &str, s: &str) -> bool {
///
/// Databricks Unity Catalog model-service names are catalog data, not model
/// family hints. Both capability interpreters use this shape check before
/// family matching so suffixes such as `kimi-k3` cannot inherit endpoint
/// capabilities accidentally.
/// family matching so services cannot inherit endpoint capabilities accidentally.
/// GPT-5+ services have a route-only Responses exception.
pub(crate) fn is_databricks_model_service_fqn(model: &str) -> bool {
let mut components = model.split('.');
let (Some(catalog), Some(schema), Some(service)) =
@@ -308,16 +308,35 @@ pub(crate) fn is_databricks_model_service_fqn(model: &str) -> bool {
}) && components.next().is_none()
}
/// Route GPT-5+ UC services to Responses without borrowing endpoint effort facts.
/// Match the first family token in the service only, preserving the existing
/// boundary semantics (e.g. `claude-gpt-5` is not a GPT service).
fn fqn_requires_responses(model: &str) -> bool {
let Some(service) = model.rsplit('.').next() else {
return false;
};
let lower = service.to_ascii_lowercase();
let stripped = strip_catalog_prefix(&lower, &manifest().family_tokens);
let Some(version) = stripped.strip_prefix("gpt-") else {
return false;
};
let digits = version.bytes().take_while(u8::is_ascii_digit).count();
let (major, suffix) = version.split_at(digits);
if suffix.starts_with(|c: char| c.is_ascii_alphanumeric()) {
return false;
}
major.parse::<u32>().is_ok_and(|major| major >= 5)
}
/// Resolve the capability profile for a `(provider, raw_model_id)` pair.
pub fn resolve(provider: &str, raw_model_id: &str) -> CapabilityResult {
let m = manifest();
let canon = canonical_provider(provider);
let blank = raw_model_id.trim().is_empty();
// Unity Catalog FQNs are neutral model-service identities. Resolve them
// through the concrete-unknown fallback before any suffix can match a
// provider family rule. Routing and effort normalization then share this
// one answer in Rust and TypeScript.
// FQNs keep neutral effort capabilities, but GPT-5+ service names need
// Responses for tools with reasoning. Only inspect the service component:
// catalog/schema names must never choose a model protocol.
let model_service_fqn =
canon == "databricks_v2" && is_databricks_model_service_fqn(raw_model_id);
@@ -394,7 +413,11 @@ pub fn resolve(provider: &str, raw_model_id: &str) -> CapabilityResult {
thinking_mode: state.thinking_mode,
supported_efforts: &state.supported_efforts,
default_effort: state.default_effort,
databricks_v2_wire_route: state.databricks_v2_wire_route,
databricks_v2_wire_route: if model_service_fqn && fqn_requires_responses(raw_model_id) {
DatabricksV2Route::OpenaiResponses
} else {
state.databricks_v2_wire_route
},
normalization_policy: state.normalization_policy,
registry_label: None,
}
@@ -724,6 +747,25 @@ mod tests {
Q::Vector { id: "boundary-claude-3-digit-run-anthropic-probe", provider: "anthropic", raw_model_id: "claude-35", note: Some("Probes whether the claude-3 prefix binds a longer digit run ('35').") },
Q::Vector { id: "boundary-claude-opus-4-70-anthropic-probe", provider: "anthropic", raw_model_id: "claude-opus-4-70", note: Some("Probes whether the claude-opus-4-7 prefix binds a longer digit run ('70').") },
Q::Vector { id: "boundary-gpt-5-1234-openai-probe", provider: "openai", raw_model_id: "gpt-5-1234", note: Some("Probes a 4-digit run after the gpt-5 stem.") },
Q::Section { group: "Databricks FQN GPT-5+ Responses routing", note: Some("Only the service component selects Responses; effort capabilities remain neutral.") },
Q::Vector { id: "dbv2-fqn-responses-0", provider: "databricks_v2", raw_model_id: "catalog.schema.goose-gpt-6-astra", note: None },
Q::Vector { id: "dbv2-fqn-responses-1", provider: "databricks_v2", raw_model_id: "catalog.schema.goose-gpt-5", note: None },
Q::Vector { id: "dbv2-fqn-responses-2", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-5-5", note: None },
Q::Vector { id: "dbv2-fqn-responses-3", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-10", note: None },
Q::Vector { id: "dbv2-fqn-responses-4", provider: "databricks_v2", raw_model_id: "catalog.schema.GOOSE-GPT-6-ASTRA", note: None },
Q::Vector { id: "dbv2-fqn-responses-5", provider: "databricks_v2", raw_model_id: "gpt-6.schema.other", note: None },
Q::Vector { id: "dbv2-fqn-responses-6", provider: "databricks_v2", raw_model_id: "catalog.gpt-5.other", note: None },
Q::Vector { id: "dbv2-fqn-responses-7", provider: "databricks_v2", raw_model_id: "catalog.schema.claude-gpt-6", note: None },
Q::Vector { id: "dbv2-fqn-responses-8", provider: "databricks_v2", raw_model_id: "catalog.schema.my-gpt-6-astra", note: None },
Q::Vector { id: "dbv2-fqn-responses-9", provider: "databricks_v2", raw_model_id: "catalog.schema.mygpt-6-astra", note: None },
Q::Vector { id: "dbv2-fqn-responses-10", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-4", note: None },
Q::Vector { id: "dbv2-fqn-responses-11", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-4o", note: None },
Q::Vector { id: "dbv2-fqn-responses-12", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-6x", note: None },
Q::Vector { id: "dbv2-fqn-responses-13", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-oss-120b", note: None },
Q::Vector { id: "dbv2-fqn-responses-14", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-", note: None },
Q::Vector { id: "dbv2-fqn-responses-15", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-4294967296", note: None },
Q::Vector { id: "dbv2-fqn-responses-16", provider: "databricks_v2", raw_model_id: "catalog.schema.gpt-６", note: None },
Q::Vector { id: "dbv2-fqn-responses-17", provider: "databricks_v2", raw_model_id: "catalog.schema.kimi-k3", note: None },
Q::Section { group: "Databricks UC model-family humanization probes (#6918 follow-up)", note: Some("Exact-record and UC-FQN strip probes for the Gemini/DeepSeek/GLM/Grok/Llama/Qwen/Gemma/Inkling families surfaced by UC discovery.") },
Q::Vector { id: "dbv2-gemini-3-1-flash-image-exact-record-probe", provider: "databricks_v2", raw_model_id: "databricks-gemini-3-1-flash-image", note: Some("Probes the Gemini 3.1 Flash Image endpoint record and label.") },
Q::Vector { id: "dbv2-gemini-3-5-flash-exact-record-probe", provider: "databricks_v2", raw_model_id: "databricks-gemini-3-5-flash", note: Some("Probes the Gemini 3.5 Flash endpoint record and label.") },
@@ -844,7 +886,7 @@ mod tests {
}
#[test]
fn corpus_has_exactly_140_executable_vectors() {
fn corpus_has_exactly_158_executable_vectors() {
// Locks the vector count so a silent INPUTS edit can't quietly drop
// coverage; must equal the gate in the TS harness
// (modelCapabilitiesCorpus.test.mjs).
@@ -853,7 +895,7 @@ mod tests {
.filter(|q| matches!(q, Q::Vector { .. }))
.count();
assert_eq!(
vectors, 140,
vectors, 158,
"corpus executable-vector count changed; update this gate deliberately"
);
}

63
crates/buzz-agent/tests/fqn_capabilities.rs Normal file
View File

@@ -0,0 +1,63 @@
//! UC route selection must change only the route, never effort or model identity.
use buzz_agent::model_capabilities::{resolve, DatabricksV2Route};
#[test]
fn gpt_fqn_route_preserves_neutral_capabilities() {
let fallback = resolve("databricks_v2", "catalog.schema.unknown");
for model in [
"catalog.schema.goose-gpt-6-astra",
"catalog.schema.gpt-5",
"catalog.schema.goose-gpt-5-5",
"catalog.schema.gpt-10",
"catalog.schema.GOOSE-GPT-6-ASTRA",
"catalog.schema.team-gpt-6-astra",
] {
let got = resolve(" DATABRICKS-V2 ", model);
assert_eq!(
got.databricks_v2_wire_route,
DatabricksV2Route::OpenaiResponses,
"{model}"
);
assert_eq!(got.thinking_mode, fallback.thinking_mode, "{model}");
assert_eq!(got.supported_efforts, fallback.supported_efforts, "{model}");
assert_eq!(got.default_effort, fallback.default_effort, "{model}");
assert_eq!(
got.normalization_policy, fallback.normalization_policy,
"{model}"
);
assert_eq!(got.registry_label, None, "{model}");
}
}
#[test]
fn unrelated_fqns_do_not_select_responses() {
for model in [
"gpt-6.schema.other",
"catalog.gpt-5.other",
"catalog.schema.claude-gpt-6",
"catalog.schema.mygpt-6-astra",
"catalog.schema.gpt-4",
"catalog.schema.gpt-4o",
"catalog.schema.gpt-6x",
"catalog.schema.gpt-oss-120b",
"catalog.schema.gpt-",
"catalog.schema.gpt-６",
"catalog.schema.gpt-4294967296",
"catalog.schema.kimi-k3",
] {
assert_eq!(
resolve("databricks_v2", model).databricks_v2_wire_route,
DatabricksV2Route::MlflowChat,
"{model}"
);
}
// The new rule is not an endpoint-family or cross-provider capability change.
assert_eq!(
resolve("databricks_v2", "goose-gpt-6-astra").databricks_v2_wire_route,
DatabricksV2Route::MlflowChat
);
assert_eq!(
resolve("openai", "catalog.schema.goose-gpt-6-astra").databricks_v2_wire_route,
DatabricksV2Route::NotApplicable
);
}

2
desktop/src/features/agents/AGENTS.md
View File

@@ -306,7 +306,7 @@ with a TypeScript lookup table or an id comparison in a component.
refresh only local persona/team/managed-agent caches; they must never
invalidate the remote relay directory.
17. **Databricks model discovery has one shared catalog authority.** Desktop and ACP call the shared `buzz-agent` discovery library; Desktop passes the effective merged `DATABRICKS_MODEL_FILTER` explicitly, and the library applies it to raw workspace endpoint IDs and Unity Catalog model-service FQNs after the additive union. A successful filtered-empty catalog is authoritative: it stays empty, disables switching, and never falls through to configured or known-model fallback. UC FQNs are catalog data and always use the MLflow Chat Completions route, regardless of family-looking text in their components. Global Defaults preserves the discovered model ID as the selected value while its closed trigger renders the provider-scoped display label; do not force the raw persisted ID over that label.
17. **Databricks model discovery has one shared catalog authority.** Desktop and ACP call the shared `buzz-agent` discovery library; Desktop passes the effective merged `DATABRICKS_MODEL_FILTER` explicitly, and the library applies it to raw workspace endpoint IDs and Unity Catalog model-service FQNs after the additive union. A successful filtered-empty catalog is authoritative: it stays empty, disables switching, and never falls through to configured or known-model fallback. UC FQNs retain neutral effort capabilities. A boundary-matched GPT-5-or-newer family in the service-name component selects OpenAI Responses so tools can coexist with reasoning; other FQNs use MLflow Chat Completions. Catalog/schema components never influence routing. Keep this route-only rule identical in the Rust and TypeScript capability interpreters. Global Defaults preserves the discovered model ID as the selected value while its closed trigger renders the provider-scoped display label; do not force the raw persisted ID over that label.
## Channel-only runtime controls

23
desktop/src/features/agents/ui/modelCapabilities.ts
View File

@@ -302,6 +302,19 @@ function isDatabricksModelServiceFqn(model: string): boolean {
);
}
// Mirror fqn_requires_responses: routing is the only inferred FQN capability.
function fqnRequiresResponses(model: string): boolean {
const service = model.split(".").at(-1) ?? "";
const stripped = stripCatalogPrefix(
service.toLowerCase(),
MANIFEST.family_tokens,
);
const match = /^gpt-([0-9]+)(?:$|[^a-z0-9])/.exec(stripped);
if (!match) return false;
const major = Number(match[1]);
return major >= 5 && major <= 0xffffffff;
}
/**
* Resolve the capability profile for a `(provider, rawModelId)` pair.
*
@@ -315,8 +328,8 @@ export function resolveModelCapabilities(
): CapabilityResult {
const canon = canonicalizeProvider(provider);
const blank = rawModelId.trim().length === 0;
// Unity Catalog FQNs are neutral model-service identities. Resolve them
// through the concrete-unknown fallback before suffix family matching.
// FQNs keep neutral effort capabilities; only GPT-5+ service names
// select Responses. Catalog/schema names never choose the protocol.
const modelServiceFqn =
canon === "databricks_v2" && isDatabricksModelServiceFqn(rawModelId);
@@ -370,7 +383,11 @@ export function resolveModelCapabilities(
// 3. Provider fallback (blank vs. concrete-unknown); never carries a label.
const pair = fallbackPair(canon);
const state = blank ? pair.blank : pair.concrete_unknown;
return toResult(state, state.databricks_v2_wire_route, null);
const route =
modelServiceFqn && fqnRequiresResponses(rawModelId)
? "openai-responses"
: state.databricks_v2_wire_route;
return toResult(state, route, null);
}
/** Authoritative list of known Databricks v2 model ids, sourced from the manifest. */

4
desktop/src/features/agents/ui/modelCapabilitiesCorpus.test.mjs
View File

@@ -25,10 +25,10 @@ const corpus = JSON.parse(readFileSync(fileURLToPath(corpusUrl), "utf8"));
// (`_group`) are skipped. Mirrors the Rust corpus filter.
const executable = corpus.filter((entry) => entry.expect != null);
test("corpus has exactly 140 executable vectors", () => {
test("corpus has exactly 158 executable vectors", () => {
// Locks the vector count so a silent corpus edit can't quietly drop coverage;
// must equal the gate in the Rust suite (model_capabilities.rs).
assert.equal(executable.length, 140);
assert.equal(executable.length, 158);
});
test("registry label aliases refuse an unprefixed query", () => {

364
scripts/normative-corpus.json
View File

@@ -2347,6 +2347,370 @@
"registry_label": null
}
},
{
"_group": "Databricks FQN GPT-5+ Responses routing",
"_note": "Only the service component selects Responses; effort capabilities remain neutral."
},
{
"id": "dbv2-fqn-responses-0",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.goose-gpt-6-astra",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-1",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.goose-gpt-5",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-2",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-5-5",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-3",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-10",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-4",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.GOOSE-GPT-6-ASTRA",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-5",
"provider": "databricks_v2",
"raw_model_id": "gpt-6.schema.other",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-6",
"provider": "databricks_v2",
"raw_model_id": "catalog.gpt-5.other",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-7",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.claude-gpt-6",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-8",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.my-gpt-6-astra",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "openai-responses",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-9",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.mygpt-6-astra",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-10",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-4",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-11",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-4o",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-12",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-6x",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-13",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-oss-120b",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-14",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-15",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-4294967296",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-16",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.gpt-６",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"id": "dbv2-fqn-responses-17",
"provider": "databricks_v2",
"raw_model_id": "catalog.schema.kimi-k3",
"expect": {
"thinking_mode": "none",
"supported_efforts": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"default_effort": "medium",
"databricks_v2_wire_route": "mlflow-chat",
"normalization_policy": "openai-clamp-max-to-xhigh",
"registry_label": null
}
},
{
"_group": "Databricks UC model-family humanization probes (#6918 follow-up)",
"_note": "Exact-record and UC-FQN strip probes for the Gemini/DeepSeek/GLM/Grok/Llama/Qwen/Gemma/Inkling families surfaced by UC discovery."