Composer 2 vs Qwen2.5-1M

Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.

Qwen2.5-1M is clearly ahead on the aggregate, 67 to 62. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Qwen2.5-1M's sharpest advantage is in agentic, where it averages 64.7 against 61.7. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 61.7% to 65%.

Composer 2 is also the more expensive model on tokens at $0.50 input / $2.50 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Qwen2.5-1M. That is roughly Infinityx on output cost alone. Composer 2 is the reasoning model in the pair, while Qwen2.5-1M is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Qwen2.5-1M gives you the larger context window at 1M, compared with 200K for Composer 2.

Quick Verdict

Pick Qwen2.5-1M if you want the stronger benchmark profile. Composer 2 only becomes the better choice if you want the stronger reasoning-first profile.

Agentic

Qwen2.5-1M

Composer 2

61.7

Qwen2.5-1M

64.7

61.7%
Terminal-Bench 2.0
65%
Coming soon
BrowseComp
72%
Coming soon
OSWorld-Verified
59%

Coding

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

73.7%
SWE Multilingual
Coming soon
97.2%
React Native Evals
Coming soon
Coming soon
HumanEval
76%
Coming soon
SWE-bench Verified
47%
Coming soon
LiveCodeBench
40%
Coming soon
SWE-bench Pro
49%

Multimodal & Grounded

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
MMMU-Pro
63%
Coming soon
OfficeQA Pro
75%

Reasoning

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
MuSR
79%
Coming soon
BBH
82%
Coming soon
LongBench v2
82%
Coming soon
MRCRv2
81%

Knowledge

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
MMLU
84%
Coming soon
GPQA
83%
Coming soon
SuperGPQA
81%
Coming soon
MMLU-Pro
74%
Coming soon
HLE
10%
Coming soon
FrontierScience
74%
Coming soon
SimpleQA
81%

Instruction Following

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
IFEval
84%

Multilingual

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
MGSM
81%
Coming soon
MMLU-ProX
80%

Mathematics

Coming soon

Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.

Coming soon
AIME 2023
85%
Coming soon
AIME 2024
87%
Coming soon
AIME 2025
86%
Coming soon
HMMT Feb 2023
81%
Coming soon
HMMT Feb 2024
83%
Coming soon
HMMT Feb 2025
82%
Coming soon
BRUMO 2025
84%
Coming soon
MATH-500
83%

Frequently Asked Questions

Which is better, Composer 2 or Qwen2.5-1M?

Qwen2.5-1M is ahead overall, 67 to 62. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 61.7% and 65%.

Which is better for agentic tasks, Composer 2 or Qwen2.5-1M?

Qwen2.5-1M has the edge for agentic tasks in this comparison, averaging 64.7 versus 61.7. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Last updated: March 18, 2026

Weekly LLM Benchmark Digest

Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.

Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.