Open Llama instruction model for multilingual chat, reasoning, and coding
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 0.3 | — | mirror | openrouter | Jul 23, 2026 |
| AA Coding Index | coding | 11.9 | — | mirror | openrouter | Jul 23, 2026 |
| AA Intelligence Index | intelligence | 9.3 | — | mirror | openrouter | Aug 7, 2026 |
| LMArena Text | preference | 1274 | — | crowd | lmarena | Sep 13, 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 |
|---|---|---|---|---|---|---|
| AkashML | fp8 | 131K | $0.2 · $0.52 | 98.44% | — | — |
| Cloudflare | fp8 | 24K | $0.293 · $2.25 | 96.21% | — | — |
| CoreWeave | fp16 | 128K | $0.71 · $0.71 | 99.96% | — | — |
| Crusoe | bf16 | 131K | $0.25 · $0.75 | 99.51% | — | — |
| DeepInfra | fp8 | 131K | $0.1 · $0.32 | 98.41% | — | — |
| — | 128K | $0.72 · $0.72 | 98.86% | — | — | |
| — | 128K | $0.72 · $0.72 | 99.96% | — | — | |
| Groq | — | 131K | $0.59 · $0.79 | 99.83% | — | — |
| Nebius | fp8 | 131K | $0.13 · $0.4 | 40.86% | — | — |
| Novita | bf16 | 12K | $0.135 · $0.4 | 95.65% | — | — |
| Parasail | fp8 | 131K | $0.22 · $0.5 | 99.39% | — | — |
| SambaNova | — | 131K | $0.45 · $0.9 | 99.12% | — | — |
| Together | — | 131K | $1.04 · $1.04 | 98.27% | — | — |