Largest Gemma 4 instruction model for open, self-hosted chat and reasoning
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 4.2 | — | mirror | openrouter | Sep 20, 2026 |
| AA Coding Index | coding | 43.4 | — | mirror | openrouter | Aug 7, 2026 |
| AA Intelligence Index | intelligence | 15.4 | — | mirror | openrouter | Sep 8, 2026 |
Includes index data from Artificial Analysis.
Providers
Live operational stats per hosting endpoint, via OpenRouter.
| Provider | Quant | Context | $/1M in · out | Uptime 24h | Latency | Throughput |
|---|---|---|---|---|---|---|
| Chutes | fp4 | 131K | $0.12 · $0.37 | 93.20% | — | — |
| CoreWeave | fp4 | 262K | $0.1 · $0.34 | 97.68% | — | — |
| Crusoe | bf16 | 262K | $0.14 · $0.4 | 98.02% | — | — |
| DeepInfra | fp8 | 262K | $0.13 · $0.38 | 96.84% | — | — |
| DeepInfra | fp4 | 262K | $0.09 · $0.34 | 99.10% | — | — |
| DeepInfra | fp8 | 131K | $0.27 · $0.76 | 97.95% | — | — |
| Friendli | — | 262K | $0.14 · $0.4 | 98.73% | — | — |
| ModelRun | fp4 | 262K | $0.75 · $1 | 99.93% | — | — |
| Novita | bf16 | 262K | $0.14 · $0.4 | 92.97% | — | — |
| Parasail | fp8 | 262K | $0.15 · $0.4 | 99.32% | — | — |
| SambaNova | — | 131K | $0.38 · $1.15 | 98.71% | — | — |
| SiliconFlow | fp8 | 262K | $0.75 · $1 | 95.36% | — | — |
| Together | — | 262K | $0.39 · $0.97 | — | — | — |
| Venice | fp4 | 256K | $0.12 · $0.36 | 98.52% | — | — |