Is Qwen2.5 72B Instruct a good deal?

Whether anything in this catalogue beats Qwen2.5 72B Instruct on both quality and price, and what you give up if it does. A computation on the current catalogue, not an opinion.

13 undominated of 136 · Sep 23, 2026

As of Sep 23, 2026, Qwen2.5 72B Instruct is dominated for Balanced on LMArena. GLM 5.3 Flash scores 203.0 higher and costs 36% less, with a covering envelope. 13 of 136 rated, priced standard models are undominated.

Inspect model evidence Compare differences & requirements

Current model

Qwen2.5 72B Instruct

Balanced

3 tokens in per 1 out

GLM 5.3 Flash is both better and cheaper than Qwen2.5 72B Instruct: 203.0 points higher on LMArena and 36% less per million tokens, $0.13 cheaper at this mix.

Qwen2.5 72B Instruct

LMArena Elo · higher is better

Scale starts at 1040 Elo

1268.9

Effective $/M · Balanced · lower is better

$0.370/M

GLM 5.3 Flash

LMArena Elo · higher is better

Scale starts at 1040 Elo

1471.9

Effective $/M · Balanced · lower is better

$0.237/M

Qwen2.5 72B Instruct takes text, returns up to 16,384 tokens from a 32,768-token context, and is listed by 2 sellers.

Compared against 136 rated, priced models on this lens: 27 models dominate it and give up nothing, 1 more dominates it but gives something up. Qwen2.5 72B Instruct scores 1268.9 at $0.37 per million tokens for this mix.

Envelope-safe replacements

Each row scores at least as high, costs no more, and covers this model’s context, output, modalities, tools, and reasoning. A cheaper narrower model is not listed here.

ModelLMArenaEffective $/MYou save
GLM 5.3 Flash1471.9 +203.0$0.237/M36%
Gemma 4 31B1441.7 +172.8$0.153/M59%
Hy31440.6 +171.7$0.144/M61%
Gemma 4 26B A4B 1434.5 +165.6$0.143/M61%
DeepSeek V4 Flash 04231431.8 +162.9$0.103/M72%
MiMo-V2.51427.4 +158.5$0.175/M53%
DeepSeek V3.21424.8 +155.9$0.302/M18%
DeepSeek V3.2 Exp1422.6 +153.7$0.305/M18%
Qwen3 235B A22B Instruct 25071419.8 +150.9$0.153/M59%
Qwen3 Next 80B A3B Instruct1417.6 +148.7$0.343/M7%
Step 3.5 Flash1403.7 +134.8$0.150/M59%
Qwen3.5-Flash1397.7 +128.8$0.114/M69%
GLM 4.5 Air1383.7 +114.8$0.310/M16%
Qwen3 30B A3B Instruct 25071383.6 +114.7$0.084/M77%
Solar Pro 41377.3 +108.4$0.158/M57%
gpt-oss-120b1365.9 +97.0$0.262/M29%
Gemma 3 27B1357.7 +88.8$0.172/M53%
GLM 4.7 Flash1352.2 +83.3$0.145/M61%
Qwen3 32B1340.0 +71.1$0.130/M65%
Gemma 3 12B1334.2 +65.3$0.075/M80%
GPT-5 Nano1319.7 +50.8$0.138/M63%
Qwen3 30B A3B1317.0 +48.1$0.215/M42%
Granite 4.2 8B1316.8 +47.9$0.107/M71%
gpt-oss-20b1287.3 +18.4$0.036/M90%
GPT-4o-mini (2024-07-18)1286.3 +17.4$0.262/M29%
GPT-4.1 Nano1284.7 +15.8$0.175/M53%
Llama 3.3 70B Instruct1274.0 +5.1$0.155/M58%

Higher score, lower price, named losses

Not a drop-in. The loss is why these are not a recommendation.

ModelLMArenaEffective $/MYou give up
Gemma 3 4B1290.7$0.063/Mtools
GLM 5.3 Flash1471.9 · $0.237/MQwen2.5 72B Instruct1268.9 · $0.370/MLMArena EloEffective $/M · Balanced
136 rated, priced standard models. Chartreuse is the frontier. Cobalt is the model you named, when it is not on the staircase.

Open the frontier

The link keeps the model and mix, using the current catalogue. Monthly spend and switching cost stay in this browser tab and are left out of the link.

Saved decision references

Save the model, workload, catalogue date and capability-preservation rule in this browser. Spend, switching cost and bill contents are not saved. No account or notifications.

Constraint: replacements must preserve the model’s capabilities; any losses remain named trade-offs.

Historical decisions cannot be fully replayed from saved references: past prices, scores and capability evidence are not stored.

Evidence & Ask