Skip to content

Commit b137299

Browse files
committed
fix(inference): keep public NVIDIA endpoint
1 parent 81512ef commit b137299

40 files changed

Lines changed: 96 additions & 96 deletions

‎agents/hermes/policy-additions.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -65,7 +65,7 @@ network_policies:
6565
nvidia:
6666
name: nvidia
6767
endpoints:
68-
- host: inference-api.nvidia.com
68+
- host: integrate.api.nvidia.com
6969
port: 443
7070
protocol: rest
7171
enforcement: enforce

‎agents/hermes/policy-permissive.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -43,7 +43,7 @@ network_policies:
4343
nvidia:
4444
name: nvidia
4545
endpoints:
46-
- host: inference-api.nvidia.com
46+
- host: integrate.api.nvidia.com
4747
port: 443
4848
protocol: rest
4949
enforcement: enforce

‎agents/openclaw/policy-permissive.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -43,7 +43,7 @@ network_policies:
4343
nvidia:
4444
name: nvidia
4545
endpoints:
46-
- host: inference-api.nvidia.com
46+
- host: integrate.api.nvidia.com
4747
port: 443
4848
protocol: rest
4949
enforcement: enforce

‎docs/inference/inference-options.mdx‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -80,7 +80,7 @@ NVIDIA Nemotron models expose OpenAI-compatible APIs across every supported depl
8080

8181
| Nemotron Host | Onboard Wizard Option | Why |
8282
|---|---|---|
83-
| `build.nvidia.com` (NVIDIA-hosted) | **Option 1: NVIDIA Endpoints** | NemoClaw sets the base URL to `https://inference-api.nvidia.com/v1` for you and validates the model against the build catalog. |
83+
| `build.nvidia.com` (NVIDIA-hosted) | **Option 1: NVIDIA Endpoints** | NemoClaw sets the base URL to `https://integrate.api.nvidia.com/v1` for you and validates the model against the build catalog. |
8484
| Self-hosted NIM container | **Option 3: Other OpenAI-compatible endpoint** | NIM exposes an OpenAI-compatible `/v1/chat/completions` route. Point the base URL at your NIM service and enter the Nemotron model ID. |
8585
| Enterprise NVIDIA AI Enterprise gateway | **Option 3: Other OpenAI-compatible endpoint** | Enterprise gateways front Nemotron with the same OpenAI-compatible contract. Use the gateway's base URL and your enterprise token. |
8686
| vLLM, SGLang, or TRT-LLM serving Nemotron weights | **Option 3: Other OpenAI-compatible endpoint** | Each runtime exposes Nemotron through `/v1/chat/completions`. Use the runtime's base URL and the model ID it reports. |
@@ -109,12 +109,12 @@ models:
109109
- name: nano
110110
litellm_model: "openai/nvidia/nvidia/Nemotron-3-Nano-30B-A3B"
111111
cost_per_m_input_tokens: 0.05
112-
api_base: "https://inference-api.nvidia.com"
112+
api_base: "https://integrate.api.nvidia.com"
113113

114114
- name: super
115115
litellm_model: "openai/nvidia/nvidia/nemotron-3-super-v3"
116116
cost_per_m_input_tokens: 0.10
117-
api_base: "https://inference-api.nvidia.com"
117+
api_base: "https://integrate.api.nvidia.com"
118118
```
119119
120120
The `tolerance` parameter controls the accuracy-cost tradeoff.

‎docs/reference/network-policies.mdx‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -41,7 +41,7 @@ The following endpoint groups are allowed by default:
4141

4242
| Policy | Endpoints | Binaries | Rules |
4343
| --- | --- | --- | --- |
44-
| `nvidia` | `inference-api.nvidia.com:443` | `/usr/local/bin/openclaw` | POST to inference and embedding paths, GET to model listings |
44+
| `nvidia` | `integrate.api.nvidia.com:443` | `/usr/local/bin/openclaw` | POST to inference and embedding paths, GET to model listings |
4545
| `clawhub` | `clawhub.ai:443` | `/usr/local/bin/openclaw`, `/usr/local/bin/node` | GET, POST |
4646
| `openclaw_api` | `openclaw.ai:443` | `/usr/local/bin/openclaw`, `/usr/local/bin/node` | GET, POST |
4747
| `openclaw_docs` | `docs.openclaw.ai:443` | `/usr/local/bin/openclaw` | GET only |

‎docs/reference/troubleshooting.mdx‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -182,13 +182,13 @@ When the lookup returns an answer, retry onboarding.
182182
### Host DNS resolution is blocked before provider validation
183183

184184
NemoClaw also checks that the host process can resolve the provider host before it starts NVIDIA provider validation.
185-
A firewall rule that blocks host DNS traffic on port `53` can make later validation fail with `curl: (6) Could not resolve host: inference-api.nvidia.com` even when container DNS probes look healthy.
185+
A firewall rule that blocks host DNS traffic on port `53` can make later validation fail with `curl: (6) Could not resolve host: integrate.api.nvidia.com` even when container DNS probes look healthy.
186186
Current onboarding stops earlier with a host DNS diagnostic and remediation hints.
187187

188188
Verify host DNS outside NemoClaw:
189189

190190
```bash
191-
node -e 'require("node:dns").resolve4("inference-api.nvidia.com", (err, addrs) => { if (err) { console.error(err); process.exit(1); } console.log(addrs.join(",")); })'
191+
node -e 'require("node:dns").resolve4("integrate.api.nvidia.com", (err, addrs) => { if (err) { console.error(err); process.exit(1); } console.log(addrs.join(",")); })'
192192
```
193193

194194
Fix the host firewall, VPN, or DNS policy so the host can resolve the provider endpoint, then rerun onboarding.

‎docs/security/best-practices.mdx‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -129,7 +129,7 @@ Endpoint rules restrict allowed HTTP methods and URL paths.
129129

130130
| Aspect | Detail |
131131
|---|---|
132-
| Default | Some endpoints allow GET and POST on `/**` (for example, `clawhub.ai`). Others restrict methods and paths to specific API routes (for example, `inference-api.nvidia.com` allows POST only to inference and embedding paths and GET to model listings). Read-only endpoints such as `docs.openclaw.ai`, the `npm_registry` baseline entry, and the `pypi` preset allow GET only (PyPI also allows HEAD). The `npm` preset is an intentional exception: npm/Yarn registry traffic uses L4 pass-through for Node 22 undici CONNECT compatibility. |
132+
| Default | Some endpoints allow GET and POST on `/**` (for example, `clawhub.ai`). Others restrict methods and paths to specific API routes (for example, `integrate.api.nvidia.com` allows POST only to inference and embedding paths and GET to model listings). Read-only endpoints such as `docs.openclaw.ai`, the `npm_registry` baseline entry, and the `pypi` preset allow GET only (PyPI also allows HEAD). The `npm` preset is an intentional exception: npm/Yarn registry traffic uses L4 pass-through for Node 22 undici CONNECT compatibility. |
133133
| What you can change | Add methods (PUT, DELETE, PATCH) or restrict paths to specific prefixes. |
134134
| Risk if relaxed | Allowing all methods on an API endpoint gives the agent write and delete access. For example, allowing DELETE on `api.github.com` lets the agent delete repositories. |
135135
| Recommendation | Use GET-only rules for endpoints that the agent only reads. Add write methods only for endpoints where the agent must create or modify resources. Restrict paths to specific API routes when possible. |

‎nemoclaw-blueprint/blueprint.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -54,7 +54,7 @@ components:
5454
default:
5555
provider_type: "nvidia"
5656
provider_name: "nvidia-inference"
57-
endpoint: "https://inference-api.nvidia.com/v1"
57+
endpoint: "https://integrate.api.nvidia.com/v1"
5858
model: "nvidia/nemotron-3-super-120b-a12b"
5959

6060
ncp:

‎nemoclaw-blueprint/policies/openclaw-sandbox-permissive.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -49,7 +49,7 @@ network_policies:
4949
nvidia:
5050
name: nvidia
5151
endpoints:
52-
- host: inference-api.nvidia.com
52+
- host: integrate.api.nvidia.com
5353
port: 443
5454
protocol: rest
5555
enforcement: enforce

‎nemoclaw-blueprint/policies/openclaw-sandbox.yaml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -73,7 +73,7 @@ network_policies:
7373
nvidia:
7474
name: nvidia
7575
endpoints:
76-
- host: inference-api.nvidia.com
76+
- host: integrate.api.nvidia.com
7777
port: 443
7878
protocol: rest
7979
enforcement: enforce

0 commit comments

Comments
 (0)