mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-05-18 04:41:56 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 7

feat(nvidia): add NIM billing origin header

Browse source
This commit is contained in:
kchantharuan 2026-05-13 12:46:07 -07:00 • committed by Teknium
parent 4e89c53082
commit 13c3d4b4ef
5 changed files with 162 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

25
agent/auxiliary_client.py
View file

@ -369,6 +369,21 @@ def build_or_headers(or_config: dict | None = None) -> dict:
return headers
# NVIDIA NIM cloud billing attribution. Keep this host-gated because the
# nvidia provider also supports local/on-prem NIM endpoints via NVIDIA_BASE_URL.
_NVIDIA_NIM_CLOUD_HEADERS = {
"X-BILLING-INVOKE-ORIGIN": "HermesAgent",
}
def build_nvidia_nim_headers(base_url: str | None) -> dict:
"""Return NVIDIA NIM cloud attribution headers for build.nvidia.com traffic."""
if base_url_host_matches(str(base_url or ""), "integrate.api.nvidia.com"):
return dict(_NVIDIA_NIM_CLOUD_HEADERS)
return {}
# Vercel AI Gateway app attribution headers. HTTP-Referer maps to
# referrerUrl and X-Title maps to appName in the gateway's analytics.
from hermes_cli import __version__ as _HERMES_VERSION
@ -1372,6 +1387,8 @@ def _resolve_api_key_provider() -> Tuple[Optional[OpenAI], Optional[str]]:
from hermes_cli.models import copilot_default_headers
extra["default_headers"] = copilot_default_headers()
elif base_url_host_matches(base_url, "integrate.api.nvidia.com"):
extra["default_headers"] = build_nvidia_nim_headers(base_url)
else:
try:
from providers import get_provider_profile as _gpf_aux
@ -1407,6 +1424,8 @@ def _resolve_api_key_provider() -> Tuple[Optional[OpenAI], Optional[str]]:
from hermes_cli.models import copilot_default_headers
extra["default_headers"] = copilot_default_headers()
elif base_url_host_matches(base_url, "integrate.api.nvidia.com"):
extra["default_headers"] = build_nvidia_nim_headers(base_url)
else:
try:
from providers import get_provider_profile as _gpf_aux2
@ -2690,6 +2709,8 @@ def _to_async_client(sync_client, model: str, is_vision: bool = False):
)
elif base_url_host_matches(sync_base_url, "api.kimi.com"):
async_kwargs["default_headers"] = {"User-Agent": "claude-code/0.1.0"}
elif base_url_host_matches(sync_base_url, "integrate.api.nvidia.com"):
async_kwargs["default_headers"] = build_nvidia_nim_headers(sync_base_url)
else:
# Fall back to profile.default_headers for providers that declare
# client-level headers on their ProviderProfile (e.g. attribution
@ -2951,6 +2972,8 @@ def resolve_provider_client(
extra["default_headers"] = copilot_request_headers(
is_agent_turn=True, is_vision=is_vision
)
elif base_url_host_matches(custom_base, "integrate.api.nvidia.com"):
extra["default_headers"] = build_nvidia_nim_headers(custom_base)
else:
# Fall back to profile.default_headers for providers that
# declare client-level attribution headers on their profile.
@ -3149,6 +3172,8 @@ def resolve_provider_client(
headers.update(copilot_request_headers(
is_agent_turn=True, is_vision=is_vision
))
elif base_url_host_matches(base_url, "integrate.api.nvidia.com"):
headers.update(build_nvidia_nim_headers(base_url))
else:
# Fall back to profile.default_headers for providers that declare
# client-level attribution headers on their profile (e.g. GMI