Models/Qwen2.5 Turbo

MDL

Qwen2.5 Turbo

Alibaba

Verified against Artificial Analysis · Jul 13, 2026

Capability

percentile index

Price /M

$0.05 / $0.2

in / out

Context

—

Avg score

36.8

6 benchmarks

AA Intelligence

6.3

index

Speed

112 tok/s

TTFT 1.1s

Released

Nov 18, 2024

About

Qwen2.5 Turbo is a open-weights model in the Qwen2.5 family from Alibaba. Benchmarked on 6 evals, averaging 36.8.

Benchmark scores · 6

avg 36.8 — higher is better

MATHartificial-analysis · 2026-07

80.5

MMLU-Proartificial-analysis · 2026-07

63.3

GPQA Diamondartificial-analysis · 2026-07

LiveCodeBenchartificial-analysis · 2026-07

16.3

SciCodeartificial-analysis · 2026-07

15.3

Humanity's Last Examartificial-analysis · 2026-07

4.2

The research behind this model

all papers mentioning it →

Qwen2.5 Technical Report71 cites · Dec 19, 2024 Qwen2.5-1M Technical Report8 cites · Jan 26, 2025

Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.

Details

PublisherAlibaba

FamilyQwen2.5

Parameters—

License—

WeightsOpen

ReleasedNov 18, 2024

Performance & price

Output speed112 tok/s

Latency (TTFT)1.1s

Price in / out$0.05 / $0.2 per 1M

Blended price$0.088 / 1M

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

History

full ledger →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$0.2/M