Is MiMo-V2.5 a good deal?

Whether anything in this catalogue beats MiMo-V2.5 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, Gemma 4 31B scores higher and costs less than MiMo-V2.5 for Balanced on LMArena, but it is not a drop-in. You would give up: context, max output, audio.

Inspect model evidence Compare differences & requirements

Current model

MiMo-V2.5

Balanced

3 tokens in per 1 out

Gemma 4 31B scores 14.3 points higher and costs 13% less, but you would give up: context, max output, audio.

MiMo-V2.5

LMArena Elo · higher is better

Scale starts at 1040 Elo

1427.4

Effective $/M · Balanced · lower is better

$0.175/M

Gemma 4 31B

LMArena Elo · higher is better

Scale starts at 1040 Elo

1441.7

Effective $/M · Balanced · lower is better

$0.153/M

MiMo-V2.5 takes text, audio, image, video, returns up to 131,072 tokens from a 1,050,000-token context, and is listed by 6 sellers.

Compared against 136 rated, priced models on this lens: nothing dominates it outright, 4 more dominate it but give something up. MiMo-V2.5 scores 1427.4 at $0.18 per million tokens for this mix.

Higher score, lower price, named losses

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

ModelLMArenaEffective $/MYou give up
Gemma 4 31B1441.7$0.153/Mcontext, max output, audio
Hy31440.6$0.144/Mcontext, max output, audio, image, video
Gemma 4 26B A4B 1434.5$0.143/Mcontext, audio
DeepSeek V4 Flash 04231431.8$0.103/Mcontext, audio, image, video
MiMo-V2.51427.4 · $0.175/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