DeepInfra

GLM-5.1

Made by Z.ai (Zhipu AI). Released 2026-04-07. Accessibility not recorded. Cheapest standard rate we hold: $3.50 per million output tokens, no context split.

What it costs

TierKindInOutCached inFrom
standardtext$1.05$3.50$0.20registry

per million tokens, US dollars. * marks a rate from a single registry or two that disagree — the best available number rather than a confirmed one. Every other row is confirmed against a seller or a licensed index.

What it scores, by house

BenchmarkHouseScoreEvaluatedScaffold
chess puzzlesEpoch AI0.2±0.042026-08-10
frontiermathEpoch AI0.3±0.032026-05-11
frontiermath tier 4Epoch AI0.1±0.052026-05-12
frontiermath tiers 1 3 v2Epoch AI0.4±0.032026-08-29
frontiermath tiers 1 3 v2Epoch AI0.2±0.032026-08-28none
gpqa diamondEpoch AI0.9±0.022026-08-10
otis mock aime 2024 2025near its ceilingEpoch AI0.9±0.042026-08-10
simpleqa verifiedEpoch AI0.3±0.012026-08-27
swe bench verifiedEpoch AI0.7±0.022026-05-15
ale benchcollected by Epoch AI887
critptcollected by Epoch AI0.0
exploitbenchcollected by Epoch AI0.2
gbaevalcollected by Epoch AI0.0
proofbenchcollected by Epoch AI0.2
scicodecollected by Epoch AI0.4
simplebenchcollected by Epoch AI0.6
vending bench 2collected by Epoch AI5634
webdev arenacollected by Epoch AI1509
weirdmlcollected by Epoch AI0.6

These scores are not comparable between houses and are not added up. A score is dated by when it was evaluated, not when the model was released, and a score without its scaffold is not a measurement — which is why both are printed.

Every rate we hold · Why there is no single best model · Everything from DeepInfra