NEWMEN
← back to all models

Google: Gemma 4 31B

google/gemma-4-31b-it

Googleopen-weight262K context11 providersIntelligence 65.0· est.

Cheapest provider

$0.12 / 1M

DeepInfra

Fastest provider (p95)

No throughput data yet — populated as traffic accumulates

Intelligence (estimated)

65.0

Family + generation + popularity + price tier

Per-provider performance

Latency / throughput / uptime / price measured across providers over the last 30 minutes of live traffic. This is what proves “sourced cheapest” — Atlas mode draws on these per call to serve the cheapest path that holds quality.

ProviderQuantInput $/1MOutput $/1MLatency p50 / p95Throughput p50 / p95Uptime 30mSuccess
DeepInfraq4· fp4$0.1200$0.370097.55%
DeepInfraq8· fp8$0.1300$0.380099.62%
SiliconFlowq8· fp8$0.1300$0.40009.94%
Novitafull· bf16$0.1400$0.400099.32%
Chutesq4· fp4$0.1500$0.420099.49%
Parasailq8· fp8$0.1500$0.400098.40%
Phalaundisclosed$0.1500$0.460099.37%
Venicefull· bf16$0.1550$0.440099.33%
Togetherundisclosed$0.2800$0.8600100.00%
WandBfull· bf16$0.3000$1.2500100.00%
Togetherundisclosed$0.3900$0.970099.59%

“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).

Intelligence estimate

No public benchmark numbers indexed for this model yet, so the leaderboard score is derived from the catalogue data we sync: model family standing, generation, and live usage rank — applied identically to open- and closed-weight models. The heuristic ceiling sits below confirmed-benchmark frontier models so curated rankings stay clearly on top.

Want a curated score? File a benchmark report and we’ll add it to the next sync.

How Atlas mode sources Google: Gemma 4 31B

  • Strict mode — pin Google: Gemma 4 31B exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently DeepInfra at $0.12/1M.
  • Atlas mode — the default. Each call is auto-optimized for the cheapest path that holds quality, at least 5% off going direct from call one and climbing as it ramps. You always see which model served the call and exactly what you saved — thumbs-down anything you don’t like for a full refund.