- Reported by
- Reported by NVIDIA in its benchmark card
- Model
- Claude Opus 4.8 (aws/anthropic/bedrock-claude-opus-4-8, as the source names it)
- Harness
- claude-code, version not stated
- Date
- 2026-10-02
- Tasks and runs
- 22 tasks, as printed
- Result
- NVIDIA reported a score of 77.9% overall and no uplift figure (1 attempt per task). Overall:
77.9% — baseline ran, but no comparable score was available; uplift unavailable
Security:100.0% → 100.0% (±0.0 points)
Correctness:29.1% → 53.6% (+24.5 points)
Discoverability:91.6% — baseline ran, but no comparable score was available; uplift unavailable
Effectiveness:39.2% → 49.1% (+9.9 points)
Efficiency:95.0% — baseline ran, but no comparable score was available; uplift unavailable
- Tokens
- 3,108,867 tokens with the skill and 1,301,936 without, over all cases, as printed; the card gives no cost
- Freshness
- Reported on Opus 4.8. The registry lists no opus model released after this run, as of the build date.
- Skill commit
- 14a98ae278cec7e867695189d0f4003c08e4f765
- Receipt
- nvidia-broadcast-aws-anthropic-bedrock-claude-opus-4-8-imported-nvidia-cards-claude-code-2026-10-02-93a8fdaeaadd.json
nvidia-broadcast: worth installing? Reported results
Reported results for nvidia-broadcast (NVIDIA/skills) from NVIDIA, one row per receipt, in plain words.
Skill skills/nvidia-broadcast in NVIDIA/skills.
One row per receipt. No figure on this page combines two rows.
- Reported by
- Reported by NVIDIA in its benchmark card
- Model
- GPT-5.5 (openai/openai/gpt-5.5, as the source names it)
- Harness
- codex, version not stated
- Date
- 2026-10-02
- Tasks and runs
- 22 tasks, as printed
- Result
- NVIDIA reported a score of 74.1% overall and no uplift figure (1 attempt per task). Overall:
74.1% — baseline ran, but no comparable score was available; uplift unavailable
Security:95.5% → 100.0% (+4.5 points)
Correctness:30.0% → 49.1% (+19.1 points)
Discoverability:86.1% — baseline ran, but no comparable score was available; uplift unavailable
Effectiveness:38.9% → 45.6% (+6.7 points)
Efficiency:89.5% — baseline ran, but no comparable score was available; uplift unavailable
- Tokens
- 1,813,207 tokens with the skill and 942,846 without, over all cases, as printed; the card gives no cost
- Freshness
- Reported on GPT-5.5. Freshness unknown: the registry has no release date for it.
- Skill commit
- 14a98ae278cec7e867695189d0f4003c08e4f765
- Receipt
- nvidia-broadcast-openai-openai-gpt-5-5-imported-nvidia-cards-codex-2026-10-02-93a8fdaeaadd.json
Figures from NVIDIA/skills at commit 14a98ae278cec7e867695189d0f4003c08e4f765, licensed CC BY 4.0. Reformatted by Driftproof; NVIDIA does not endorse this page.
Built on from committed receipts.