Benchmark Almanac

Qwen3-1.7B

Qwen3-1.7B.

Qwen Released April 29th, 2025 Compare with others
42
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Qwen Apr 2025

Qwen3-1.7B

Open weights

119.2 #350 of 404
Self-Reported Index, from the lab's own numbers. No Epoch Capabilities Index score yet.
  • GPQA Diamond 38% #190 /229
  • LiveCodeBench v5 33.2% #42 /47
  • KINA 18.3% #35 /37

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
ArabicMMLU 49.7% #17 of 39 Official leaderboard
ArbMMLU-HT 43.9% #19 of 39 Official leaderboard
Open Arabic LLM Leaderboard v2 52.9% #20 of 39 Official leaderboard
Vectara Hallucination Leaderboard (HHEM) 4.40% #77 of 148 Official leaderboard
AlGhafa 56.8% #21 of 39 Official leaderboard
MadinahQA 40% #23 of 39 Official leaderboard
ALRAGE 66.3% #24 of 39 Official leaderboard
Global PIQA 60.8% #30 of 44 Official leaderboard
MathIF 30.2% #9 of 12 Official leaderboard
Encyclo-K 18.3% #19 of 23 Official leaderboard
AlignBench 7.60 #13 of 15 Lab model card
MLogiQA 56% #7 of 8 Lab model card
MMMLU 59.1% #63 of 71 Lab model card
IFEval 72.5% #122 of 135 Lab model card
C-Eval 68.1% #42 of 46 Lab model card
Wer wird Millionär? LLM Benchmark 84.0 #71 of 77 Official leaderboard
Arena-Hard-Auto 43.1% #36 of 39 Lab model card
Multi-IF 51.2% #26 of 28 Lab model card
INCLUDE 51.8% #29 of 31 Lab model card
KINA 18.3% #35 of 37 Official leaderboard
MMLU-Redux 73.9% #43 of 45 Lab model card
WritingBench 70.2 #14 of 14 Lab model card
Creative Writing v3 48.0 #12 of 12 Lab model card
MT-AIME2024 36.1% #7 of 7 Lab model card

Math and science

BenchmarkScoreRankSource
MATH-500 93.4% #47 of 70 Lab model card
OTIS Mock AIME 2024-2025 8.06% #166 of 212 Epoch AI run
GPQA Diamond 38% #190 of 229 Epoch AI run
AIME 2024 48.3% #76 of 88 Lab model card
ZebraLogic 63.2% #21 of 23 Lab model card
AIME 2025 36.8% #135 of 143 Lab model card
PolyMath 25.2% #23 of 24 Lab model card
AutoLogi 83.2% #8 of 8 Lab model card

Coding

BenchmarkScoreRankSource
LiveCodeBench v5 33.2% #42 of 47 Lab model card
AutoCodeBench 11.2% #33 of 34 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
C3-Bench 6.25% #19 of 30 Official leaderboard
BFCL V4 28.4% #42 of 65 Official leaderboard
LisanBench 133.3 #105 of 133 Official leaderboard
Berkeley Function Calling Leaderboard (BFCL) 56.6% #39 of 41 Lab model card

Professional work

BenchmarkScoreRankSource
BRIDGE 34.8 #34 of 39 Official leaderboard
BizFinBench 50.8% #12 of 13 Official leaderboard

Safety and security

BenchmarkScoreRankSource
AraTrust 75.6% #20 of 39 Official leaderboard

Other

BenchmarkScoreRankSource
LiveBench 51.1% #19 of 27 Lab model card

Its lab's numbers put it at 119.2, number 350 of 404

The Self-Reported Index fits one score to the 17 capability results Qwen reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 117.0 to 121.4. How we compute it.

Its lab's numbers next to 3 leaderboards

Where the benchmark's own leaderboard, or Epoch AI's run, also has this model. Gaps mostly come from different harnesses, effort levels and task subsets. Every pair.

Benchmark Lab says Board says Gap
GPQA Diamond 40.1% GPQA-Diamond, thinking mode (Table 19). Non-thinking 28.6 (Table 20) 38% Epoch AI run +2.1 pts
Berkeley Function Calling Leaderboard (BFCL) 56.6% BFCL v3, thinking mode (Table 19). Non-thinking 52.2 (Table 20) 54.7% Leaderboard, native function calling +1.9 pts
PolyMath 25.2% PolyMath, thinking mode (Table 19). Non-thinking 10.3 (Table 20) 25.2% Leaderboard, thinking Same

Its system card

Every result in the longer documents the lab published with it. 23 of them aren't in the launch post.