Chatglm2 6B — Benchmarks

Benchmark scores for Chatglm2 6B aggregated from public leaderboards (8 boards), with how it ranks among open models and how it compares with its nearest neighbours. See hardware requirements for what you need to run it.

No llmrun Composite Score for Chatglm2 6B: the score is only published for a model measured on at least 4 of the boards in the score panel, spanning at least 2 benchmark categories and including a reasoning or coding board — or on 3 boards that narrow its score to a 90% range of at most 15.6 points. Either way the results must come from at least two different benchmarks. Its individual scores below stand on their own — see how the llmrun Composite Score gets calculated.

Chatglm2 6B reasoning benchmarks

Multi-step logic and problem-solving — can the model think through a hard problem rather than recall an answer.

BIG-Bench Hard
BIG-Bench Hard (BBH) is the 23-task subset of BIG-Bench that earlier models found hardest — multi-step reasoning, logic, and word problems. It's a long-standing standard for tracking reasoning ability across model generations.
33.7%
33.7%
median 50.8%open leader 87.5%
Overall best: Gemini 1.5 Pro 001 89.2%
11th percentile · #34 of 38 open · #46 of 50 all
#34 of 38
#46 of 50
Lab-reported
Epoch AI · Jun 2023

Other results (1)leaderboard-only, not part of the llmrun Score

Leaderboard-only, not part of the llmrun Score

PIQA
PIQA (Physical Interaction QA) asks a model to choose the more sensible of two ways to accomplish an everyday physical task, testing practical, physical-world commonsense that isn't well captured by purely textual reasoning tests.
69.6%
69.6%
median 80.1%open leader 88.6%
Overall best: GPT 4o Mini (Jul 18, 2024) 88.7%
3rd percentile · #39 of 40 open · #58 of 59 all
#39 of 40
#58 of 59
Lab-reported
Epoch AI · Jun 2023

Chatglm2 6B math benchmarks

Symbolic and quantitative problem-solving, from competition math to multi-step calculation.

GSM8K
GSM8K is a classic set of grade-school math word problems requiring several arithmetic reasoning steps. Long a default sanity check for reasoning, top models now score near the ceiling, so it best separates smaller models.
32.4%
32.4%
median 52.3%open leader 94.5%
26th percentile · #46 of 62 open · #64 of 93 all
#46 of 62
#64 of 93
Lab-reported
Epoch AI · Jun 2023

Chatglm2 6B knowledge benchmarks

Breadth and depth of factual knowledge across many subjects, paired with the reasoning to apply it.

MMLU
MMLU (Massive Multitask Language Understanding) spans 57 subjects from history to law to medicine as multiple-choice questions. It is the long-standing default for broad knowledge and remains the most widely-reported general benchmark.
47.9%
47.9%
median 62.5%open leader 87.2%
Overall best: GPT 4o (Nov 20, 2024) 88.1%
25th percentile · #64 of 85 open · #111 of 136 all
#64 of 85
#111 of 136
Lab-reported
Epoch AI · Jun 2023

Other results (3)leaderboard-only, not part of the llmrun Score

Leaderboard-only, not part of the llmrun Score

ARC Challenge
ARC Challenge is a set of grade-school science questions, curated by the Allen Institute for AI (AI2) to exclude ones that simple word-matching or retrieval can answer. It measures science knowledge paired with the reasoning needed to apply it.
61.0%
61.0%
median 53.9%open leader 95.3%
60th percentile · #23 of 56 open · #33 of 76 all
#23 of 56
#33 of 76
Lab-reported
Epoch AI · Jun 2023
HellaSwag
HellaSwag is a commonsense benchmark that asks a model to pick the most plausible continuation of an everyday situation. The wrong options are adversarially chosen to fool models, testing grounded commonsense understanding.
57.0%
57.0%
median 76.2%open leader 89.2%
Overall best: GPT 4 (Mar 14) 95.3%
13th percentile · #41 of 47 open · #65 of 75 all
#41 of 47
#65 of 75
Lab-reported
Epoch AI · Jun 2023
BoolQ
BoolQ is a reading-comprehension benchmark of naturally occurring yes/no questions, each paired with a short passage that contains the answer. It tests whether a model can extract and reason over information from real text rather than just recall facts.
79.0%
79.0%
median 76.4%open leader 89.4%
Overall best: T5 11B 91.2%
59th percentile · #16 of 38 open · #41 of 76 all
#16 of 38
#41 of 76
Lab-reported
Epoch AI · Jun 2023

Chatglm2 6B instruction following benchmarks

How precisely a model obeys explicit, verifiable instructions — format, length, and content constraints — and, more broadly, how much people prefer its responses head-to-head in blind human voting.

LMArena Text
Arena (arena.ai, formerly LMArena / Chatbot Arena) ranks models by millions of blind, head-to-head human votes on real prompts. The Arena Score shown here is a Bradley-Terry rating computed with Arena's default style control, so higher means people preferred that model's answers more often — it's a relative rating, not a percentage, so compare it only to other models on this board.
1024.8
1024.8
median 1321.4open leader 1488.2
Overall best: Gemini 4 Carbon (High) 1527.5
1st percentile · #187 of 188 open · #402 of 408 all
#187 of 188
#402 of 408
Arena
Arena · Oct 2026

Scores aggregated from public benchmark sources (Epoch AI: 7 boards, Arena: 1 board) — each linked from the benchmark pages. llmrun does not run these benchmarks. Percentiles and ranks are among open models in the llmrun catalogue; the median and leader marks refer to the same field.