Description

When a model ID contains a prefix that matches the provider name (e.g., nvidia/nemotron-3-super-120b-a12b on NVIDIA provider), OpenClaw incorrectly strips the prefix when making API requests, causing 404 errors.

Steps to reproduce

1. Add model with ID nvidia/nemotron-3-super-120b-a12b to NVIDIA provider config
2. Reference it in fallbacks as nvidia/nvidia/nemotron-3-super-120b-a12b
3. Trigger fallback to this model

Expected behavior

API receives model name: nvidia/nemotron-3-super-120b-a12b

Actual behavior

API receives: nemotron-3-super-120b-a12b → 404 Not Found

Evidence

Direct curl test to NVIDIA API:

# Works ✅
curl https://integrate.api.nvidia.com/v1/chat/completions \
  -d '{"model": "nvidia/nemotron-3-super-120b-a12b", ...}'

# Fails ❌
curl https://integrate.api.nvidia.com/v1/chat/completions \
  -d '{"model": "nemotron-3-super-120b-a12b", ...}'
# Returns: 404 page not found

Log shows OpenClaw sends wrong model name:

{"model":"nemotron-3-super-120b-a12b","provider":"nvidia","rawErrorPreview":"404 404 page not found"}

Other NVIDIA models work fine

• z-ai/glm5 ✅
• moonshotai/kimi-k2.5 ✅
• deepseek-ai/deepseek-v3.2 ✅

Environment

• OpenClaw version: 2026.3.13
• Provider: NVIDIA (integrate.api.nvidia.com)