Meta 🇺🇸 · Llama 3.3

Llama 3.3 70B Instruct

An open-weight 70-billion-parameter instruction model approaching Llama 3.1 405B performance at substantially lower serving cost with multilingual and tool-use support.

DOOM SCORE67.7out of 100model risk profile, not the overall index
0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 3

Why this model scores 67.7

Capability improves modestly over Llama 3.1 70B and approaches the prior 405B tier, while autonomy remains application-mediated. Open weights, lower cost and broad hosting create near-maximal deployment, misuse availability and control difficulty.

Capability75
Autonomy22
Deployment98
Misuse potential72
Control difficulty76
MODEL-ATTRIBUTED EVIDENCE

News tied to Llama 3.3 70B Instruct

The model score of 67.7 rates this model's risk profile. The overall Doom Index of 67.9 measures the complete temporally weighted evidence record. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.10

Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.

Capability TOWARD

Meta releases Llama 3.3 70B Instruct

Meta released Llama 3.3 70B Instruct with performance approaching Llama 3.1 405B Instruct at much lower serving cost under the Llama community license.

Full item contribution
+0.10
Llama 3.3 70B Instruct equal share
+0.10
Read assessment →
AUDIT TRAIL

Model score history

  1. R3
    Doom Score 67.7

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-142416 under temporal-monthly-pressure-v4.

    12 Aug 2026
  2. R2
    Doom Score 68.6

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-132112 under fixed-sensitivity-v3.

    12 Aug 2026
  3. R1
    Doom Score 70.9

    Initial exact source-backed profile for Llama 3.3 70B Instruct.

    12 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for Llama 3.3 70B Instruct.
  1. Llama 3.3 70B Instruct by Meta has a DoomBench model risk score of 67.7 out of 100, based on five transparent version-specific dimensions rather than the overall index.

  2. Llama 3.3 70B Instruct's highest current DoomBench dimension is deployment at 98.0 out of 100; the profile publishes every component score and its editorial rationale.

  3. DoomBench links 1 source-backed evidence item to Llama 3.3 70B Instruct, while keeping the model's risk profile separate from each item's contribution to the live Doom Index.

    https://www.doombench.com/models/meta-llama-3-3-70b-instruct